DEV Community: Viktor Kondas

A hands-on Supabase + Vercel alternative: Lathe’s one-box stack

Viktor Kondas — Wed, 04 Nov 2026 09:02:20 +0000

Disclosure: This is a sponsored hands-on review. Lathe compensated me for the write-up. Opinions, timings, and “who it’s for” notes are from my own Ship Notes test on Free and Starter. Links to Lathe are product links from that review — not affiliate / referral links.

I wanted one place to run a small product: the app, Postgres, and sign-in — without juggling five dashboards on day one.

So I tested Lathe with a throwaway app I called Ship Notes — Lathe Free, then Starter. First Apps + Postgres + Auth. Then the full engine set: Redis, SeaweedFS, NATS, and CouchDB. I timed a Free cold wake. I restored a daily backup and undid it. I left notes over a weekend.

This is what I clicked, what came back, and who I think Lathe is for.

Try the live demo: Ship Notes on Lathe Starter — create an account with email + password (sign-ups are open). See all seven engines on /engines. Sign up for your own box: lathe.live.

What Lathe is (in one breath)

Lathe gives you one dedicated machine per instance. You pick the engines that fit the plan’s RAM: your container app, Postgres 17, Auth built on the open-source Supabase Auth server, plus optional Redis 8, CouchDB 3, NATS 2, and SeaweedFS 4 (S3-compatible). Lathe runs the machine — TLS, credentials, monitoring, patching, and on paid plans daily backups.

It is not “self-host Coolify and you are on-call.” It is also not “Vercel for the frontend + Supabase for the backend + another vendor for Redis.” It is one box, one bill, one operator surface (portal, REST API, or MCP for agents).

If you already like the Supabase Auth + Postgres mental model, Lathe will feel familiar — users live in your instance Postgres auth schema, and supabase-js is the client path Lathe documents.

For the wider free-stack map (hosting, databases, auth, email), I keep separate guides: free hosting, database and storage, auth, and email / SMS. This piece is the hands-on “one box instead of five tabs” review.

What I ran: Free vs Starter

	Free (as tested)	Starter (as tested)
Specs	0.5 vCPU · 1 GB · 10 GB	2 vCPU · 4 GB · 40 GB
Engines	Apps + Postgres 17 + Auth	All seven
Traffic (plan)	1000 GB/mo	20 TB/mo
Daily backups	No	Yes (7 kept)
List price	$0	$29/mo

Flat pricing

Mini / Plus / Pro sat at $15 / $49 / $89 — flat monthly machines, large paid traffic allotment (Lathe listed 20 TB/mo), nothing metered per request or auth MAU in the story Lathe publishes. See pricing.

Pricing screenshot mid-October 2026.

What I built: Ship Notes

Ship Notes is a small notes app: sign in, post a note, optionally attach a photo, optionally mark a note public. Enough to touch Auth, Postgres, object storage, cache, a message bus, and a document store — without pretending it is a production SaaS.

Deploy path I used: build a linux/amd64 container image, push to a public GHCR tag, point Lathe’s Apps engine at that image. Contract is simple: listen on $PORT, answer /health with 2xx, keep durable state under /data if you need it. No Lathe GitHub App required. Lathe says you can also connect GitHub and let it build from a repo; I stayed on the image path to keep this review repo isolated from production tooling.

I also drove parts of the test from an agent session against Lathe’s API / MCP-style ops: wake, deploy, restore. Destructive calls need confirm equal to the instance id. Payment stays with a human. That matched how I want infrastructure products to behave around coding agents.

Free tier: Apps + Postgres + Auth

I created a Free instance and turned on three engines. Specs I saw: 0.5 vCPU · 1 GB · 10 GB NVMe, 1000 GB traffic / month, no daily backups, one Free instance per person.

Sign-up: password works; magic link needs SMTP

When users sign up, Lathe does not ship Clerk-style magic link delivery by default.

Lathe Auth creates users, sessions, and magic-link / OTP tokens. It is the sign-in engine (open-source Supabase Auth / GoTrue on your machine). It is not a hosted mail product. Users live in your Postgres auth schema. With SMTP empty on my Free instance, magic link / OTP could return HTTP 200 while nothing hit the inbox. Auth was fine; mail transport was not set up.

Path	Needs SMTP?	What I saw
Magic link / OTP	Yes	200, no mail (and later rate-limit 429)
Password sign-up	No	Session returned immediately
Password sign-in	No	Works once the user exists

	Clerk (typical)	Lathe Auth (what I saw)
Creates magic link / OTP	Yes	Yes
Sends the email	Included	Only if you configure SMTP
Users / sessions	Their cloud	Your Postgres `auth` schema

So self-serve onboarding without SMTP is Create account with password, not magic link. That is what I used on Ship Notes — no admin bootstrap required. If you want “email me a link” out of the box, bring SMTP (Lathe’s Auth settings even hint at Resend-style smtp_user / API key) or stay with a product that includes delivery. For other auth options, see the auth guide; for transactional mail providers, see email / SMS.

One sharp edge for agents and docs: pass the Auth base URL into supabase-js (do not leave a trailing /auth/v1 if Lathe already injects that path — you will double it). Set Auth site_url to your app so a future magic link returns to the right place after SMTP is wired.

First note landed in Postgres (Lathe free tier test). More notes over the following days — five total before the idle test. After a weekend away, the data was still there. Lathe’s weekly status mail said the machine was fine (disk ~30%, no failed changes). Free does not run daily restore tests — expected with no daily backups.

I closed the abuse surface once my test user existed: app SIGNUPS_OPEN=false, Auth sign-ups closed, allow-list for posts. On the public Starter demo later I re-opened sign-ups so you can try it.

Free sleeps — and Lathe tells you

Lathe emailed me when the Free instance slept after 1440 minutes idle (24 hours — what I saw on this plan; treat idle rules as “check the portal / plan”).

I woke it via the API. First /health 200 at about 51 seconds. Portal status running at about 91 seconds. Same hostname. All five notes still there. Lathe says paid plans do not sleep on their own.

That sleep/wake loop is the clearest Free tradeoff: great for a tryout, not for something that must answer in under a second at 3am with no traffic.

Operator UX detail I liked: Lathe emails when Free sleeps, when a machine is waking / ready, and when a paid instance is ready. You are not guessing from a silent dashboard.

Starter: the full stack on one machine

I moved the same product shape onto Starter ($29/mo) — full stack label. All seven engines on: Apps, Auth, Postgres, Redis, CouchDB, NATS, SeaweedFS.

What I verified on Starter

On Free, three engines was the honest ceiling for 1 GB. On Starter, turning everything on is the point of the plan. Memory splits across engines (Lathe’s pricing page shows a full Starter allocation example; the portal chart matched the “many engines sharing one box” story).

Postgres 17 — notes persisted; pooled and direct ports available; survived image rollback.
Auth — password sign-up / sign-in; site_url set for the app; users still in the instance auth schema.
SeaweedFS — photo upload ≤2 MB on notes (S3-compatible bucket on the same machine).
CouchDB — JSON copy / mirror of notes when the app env pointed at it.
Redis — rate limit + cached note count path in the app.
NATS — publish note.created when configured.
Deploy / rollback — I bumped a release env; /health stayed HTTP 200 across 45/45 checks during the restart, then I rolled the image tag back and forward — notes and photos survived.
Guardrails — delete instance and restore both need an explicit confirm (and restore needs the image id). Good default when an agent has keys.

The public demo is sanitized: welcome note, sign-ups open, password path. That is intentional so readers can poke the box without seeing my private test content.

Backups and restore (timed)

Paid plans keep daily full-disk images (seven retained). Lathe restores in place — connection strings and hostname stay put. Before a restore, Lathe keeps a short-lived undo snapshot.

I restored a daily backup, then undid via the pre-restore snapshot Lathe took automatically. Hostname did not change either time.

Action	App `/health` non-200 window	Job done
Restore daily backup	~59.5s	~127s
Undo (restore snapshot)	~65.3s	~130s

That is not point-in-time recovery. Lathe is upfront that paid backups are daily (RPO up to about a day of writes). If you need finer recovery, stay on a provider that sells PITR. If daily images plus a tested restore button are enough for a small product, the path I ran worked.

When Lathe is a great fit

Sign up at lathe.live when your next project is small and you do not want Vercel + Neon + Clerk + storage running in parallel on day one.

I would reach for Lathe first when:

A small product or side project — one box for app, DB, sign-in instead of five vendor tabs.
You already think in Supabase-style Auth + Postgres and want that on your own VM, with users in your database and a flat monthly bill.
Save the parallel stack bill: Vercel Pro + Supabase Pro + Clerk Pro was about $70/mo vs Lathe Starter ($29/mo) for app + Postgres + Auth (and the other engines) on one invoice.
You are fine with one machine (grow by upgrading the plan, not by adding replicas).
You want agents in the loop: Lathe exposes portal, API, and MCP, with destructive actions gated behind confirmation and payment left to a human.
Lathe says EU regions are available if you care; I did not make residency the story of this review — the product story for me was operational simplicity.

Start on Free if you only need Apps + Postgres + Auth and can live with sleep + no daily backups. Move to Starter ($29/mo) when you want Redis, object storage, queues, document store, always-on, and daily backups on the same box.

When I would stay on Neon, Vercel, Clerk, and friends

Lathe is a warm recommendation for the right size of problem. I would keep (or start on) the large managed stack when:

You need global edge, automatic horizontal scale, or HA / failover / read replicas.
You need point-in-time recovery or a stricter RPO than a daily image.
Magic-link / OTP mail must work with zero SMTP setup — use something like Clerk (auth guide), or wire SMTP on Lathe and own delivery (email guide).
The app is already a serious multi-service production system. Do not migrate a large live product because a review was positive. Try Lathe on a new small project or staging first.
You want Vercel preview deployments, Neon branching, or a wide serverless ecosystem more than one-box simplicity.

Other hosting and database options in the free-stack guides: hosting, database and storage.

Lathe’s own limits page is worth reading before you buy: one machine is the feature and the ceiling. Support is founder email, not an enterprise desk. That matched the product I tested — small team, honest scope.

What surprised me

Password onboarding without SMTP was clean — I expected more friction; sign-up returned a session immediately.
Status email when Free slept — small thing, big trust. Same for ready / wake mails.
Restore + undo with the same hostname — felt like a real operator tool, not a marketing checkbox.
Magic link silence without SMTP — easy to misread as “Auth is broken” if you only test OTP. It is not; magic link stays silent until you configure SMTP (Resend-style settings work in Auth).
Agent-shaped guardrails — confirm tokens on delete/restore — made me more willing to let an agent hold an API key.

Recommendation

For the next small product where you want Supabase-style Auth and Postgres next to the app — try Lathe. Free is enough to feel the shape. Starter ($29/mo) is where the one-box story gets real: backups, storage, cache, messaging, documents, always-on.

I would not rip a large multi-vendor production app onto Lathe because a review was warm. I would open lathe.live, spin Free or Starter, deploy a container, and ship something small on one machine. If the flat bill and the ops emails feel right, stay. If you outgrow one VM, you already know the exit ramps — standard Postgres, standard containers, Auth schema you can reason about.

Live demo (sign up with password): Ship Notes · all engines

Get a machine: lathe.live

Sponsored by Lathe. This post was originally published on Vibe Coder's Life (canonical URL set in post settings).

Which AI coding tool should you pay $20 for?

Viktor Kondas — Fri, 23 Oct 2026 11:00:00 +0000

Cursor vs Claude Code vs Codex vs Antigravity — a fit guide for vibe coders on one paid tier. Full roster, no single winner. (Interactive 5-question chooser lives on the site.)

Short answer: there is no universal best $20 AI coding plan. The right pick depends on where you live (IDE vs terminal), your cloud stack, and whether you work in steady daily sessions or bursty agent runs. This post is for vibe coders choosing a first or only paid tier — not for people already on $100–$200 Max plans.

Facebook and Reddit groups ask the same fight every week: Codex or Claude? Cursor or Antigravity? Use the decision map below, then read the roster. Prices are US list prices re-checked mid-October 2026 (including Cursor’s new ladder and Gemini 4 Argon). Plans change monthly — confirm on the vendor page before you pay.

What I pay today: Cursor Pro (legacy $20) and Claude Pro / Claude Code ($20). New Cursor buyers now see Plus $8 / Premium $30 / Super $100 / Ultra $200 — Premium is the serious IDE step above a strict $20 band. I spent a free Plus month inside Codex. I am in the AWS AI CEE Challenge and will try Kiro CLI next. Full stack: My vibe coding stack in 2026.

Interactive chooser (on the site)

DEV can’t run the quiz JS. Open the 5-question chooser on the original post — five questions, four answers each, you get two recommended fits (not a fake #1):

→ Open the $20 chooser

Static decision map (same 5 questions)

Question	If you pick…	Lean toward
Where do you want to live all day?	One AI IDE	Cursor (Premium) / Antigravity
	Terminal / CLI agent	Claude Code / Codex / Muse / OpenCode
	Current editor + plugin	Copilot
	Cloud agent / hand off tickets	Devin / Codex
What is your cloud / data home?	Google (Firebase, GCP, Gemini)	Antigravity
	AWS	Kiro
	EU-hosted / open weights	Mistral / Kimi / GLM / OpenCode
	Agnostic / mixed	Cursor / Claude / Codex
How do you work?	Steady daily coding	Cursor / Copilot
	Bursty multi-hour agent sessions	Claude / Codex / Antigravity
	Specs / tasks before code	Kiro
	Spiky — hate monthly waste	Muse / GLM / Kimi
What do you care about most?	Best multi-file agent quality	Claude / Codex
	Many models in one place	Cursor
	Published limits / predictability	Codex / Copilot / OpenCode
	Lowest entry price (~$10)	OpenCode Go / Copilot / Cursor Plus
What is already in your wallet?	Nothing yet — one $20 slot	Cursor / Claude / Codex / OpenCode
	ChatGPT Plus already	Cursor / Claude
	Claude Pro already	Cursor / Codex
	Cursor / Google / AWS credits already	Cursor / Antigravity / Kiro

How to read $20 plans

Most plans no longer sell “unlimited”. They sell a quota:

5-hour window + weekly cap — Claude Pro, ChatGPT Plus (Codex), Google AI Pro (Antigravity), Kimi, GLM. Heavy bursts lock you out for a few hours; weekly still sits on top.
Weekly / monthly pools — Cursor’s new ladder (Plus / Premium / Super / Ultra), Kiro, GitHub Copilot Pro. Steady daily use fits; burn early and you wait or pay overage.
~$8–$10 entry band — Cursor Plus ($8, Grok-oriented), Copilot Pro (editor plugin), and OpenCode Go (curated low-cost models for agents, with published 5-hour / monthly limit tables).
Request / usage plans near $20 — Muse Code High (~$15) sits in this band; Power is $50. API keys stay pay-as-you-go.

Pick the meter that matches how you work. Bursty agent weekends suit windows. Daily shipping suits a monthly pool.

Full roster at a glance

If this is you	Look at	US list	Meter
One AI IDE (new buyers)	Cursor Premium (Plus $8 light)	$30 / $8	Shared usage pool
Legacy Cursor Pro (existing)	Cursor Pro	$20/mo	Monthly pools
Frontier multi-file agent in terminal	Claude Pro / Claude Code	$20/mo	5h + weekly
Prefer OpenAI / Codex	ChatGPT Plus (Codex)	$20/mo	5h + weekly
Google stack (Firebase, GCP, Gemini)	Google AI Pro / Antigravity	~$19.99/mo	5h + weekly
AWS + specs before code	Kiro Pro	$20/mo	Monthly credits
Cheapest editor plugin	GitHub Copilot Pro	$10/mo	Monthly credits
Low-cost agent models (any harness)	OpenCode Go	$10/mo	5h + monthly tables
EU hosting / open weights	Mistral Le Chat Pro + Vibe	$14.99/mo	Plan-defined
Chinese open-weight route	Kimi / GLM Coding Plan	~$18–19/mo	5h + weekly
Hand off whole tickets	Devin Pro / Desktop	$20/mo	Daily/weekly quotas
Meta terminal agent, flat plan	Muse Code High	~$15/mo	Request plan (+ API option)

Cursor — new ladder (Plus $8 / Premium $30)

Pick it when you want a single AI IDE with agent mode and a shared usage pool — and you can stretch to Premium at $30 for the serious new-buyer plan. Cursor is still the IDE home-base pick; it just no longer lands cleanly at $20 for new checkouts.

New-buyer US list (mid-Oct 2026): Free · Plus $8 · Premium $30 · Super $100 · Ultra $200 (before VAT).
Plus ($8): light / Grok-oriented step — closer to the $10 lane than to Claude Pro.
Premium ($30): the plan Cursor recommends for new AI users — cloud agents and a larger shared usage pool. This is the real “one IDE” competitor in this guide, just $10 above a strict $20 budget.
Legacy Pro ($20): still exists for some existing subscribers (including me). Two monthly usage pools — Cursor models and Other Models — plus optional on-demand. Not the default new-buyer card anymore.
Watch out for: third-party models are often billed separately at API rates on the new ladder. Set a spend cap. Heavy agent days still burn included usage before the period resets.

→ cursor.com/pricing

Claude Pro (Claude Code) — frontier agent work

Pick it when multi-file refactors, long agentic sessions, and terminal-first workflows are your default.

What $20 gets you: Claude Code in the terminal, VS Code forks (including Cursor), and JetBrains. Chat and coding share one pool with a five-hour session limit plus a weekly cap.
Watch out for: Anthropic does not publish exact Pro quotas. A long chat day shortens your coding day.

→ claude.ai/code

ChatGPT Plus (Codex) — when you prefer OpenAI

Pick it when you like OpenAI models, want published limit ranges, or want Codex plus GitHub review / Slack in the same plan.

What $20 gets you: Codex on web, CLI, IDE extension, and iOS. Usage rides five-hour windows and weekly caps.
Hands-on: I used a free Plus month for real projects — One month with Codex on ChatGPT Plus. Strong second brain; Cursor stayed home base.

→ chatgpt.com/codex

Google AI Pro (Antigravity) — Google stack

Pick it when you build on Firebase, Google Cloud, or the Gemini API and want an agent-first IDE in the ~$20 band.

What ~$20 gets you: Antigravity agent capacity via Google AI Pro (~$19.99/mo). Quota refreshes on a rolling window up to a weekly cap. Free individual tier exists with tighter weekly limits.
Gemini 4 Argon (updated 16 Oct): Argon now works on AI Pro in the Gemini app, AI Mode, and the Antigravity desktop app and CLI — paid, non-trial Pro only, and not in the Antigravity IDE itself (the IDE extensions do get it). Ultra gets 5x or 20x the Pro usage quota. Argon in AI Studio with your plan is still Ultra-only. API access is paid-only (intro $2/$10 per 1M in/out through 30 Nov, then $4/$20).
Watch out for: free quota has been tightened more than once. Start free; upgrade when the weekly cap blocks real work. Cap API spend if you route Argon through Cursor, OpenRouter, or a gateway.

→ antigravity.google · Gemini 4 Argon announcement · Argon in Antigravity · Antigravity models by plan

Kiro — AWS + spec-driven

Pick it when your stack is AWS and you like requirements and tasks written before code.

What $20 gets you: free tier with monthly credits, bonus credits for new users, Pro at $20/mo for a larger credit pool. Models are mostly Anthropic/OpenAI through the product — not “Amazon Nova only”.
My angle: I am active in the AWS AI CEE program and will report back after deeper Kiro CLI use.

→ kiro.dev

GitHub Copilot Pro — the $10 layer

Pick it when you live in GitHub, VS Code, or JetBrains and want the cheapest serious paid plugin — or a second tool beside an agent plan.

What $10 gets you: monthly AI credits for chat/CLI/agent; inline completions are typically outside that meter.
Watch out for: heavy agent use burns credits quickly after the mid-2026 credit switch.

→ github.com/features/copilot

OpenCode Go — the other $10 lane

Pick it when you want a low-cost subscription for curated open / affordable coding models, usable with OpenCode or other agents — without jumping straight to a $20 frontier plan.

What $10 gets you: OpenCode Go at $10/month with generous limits on a curated agent-coding lineup (Kimi, GLM Flash, DeepSeek Flash, Muse Spark Contributor where available, and more). They publish estimated requests per five hours plus monthly usage caps by model. Go Plus is $40/month for higher limits.
Watch out for: this is model access for agents, not a full IDE replacement like Cursor. Top up credit is available if you burn through the plan; cancel anytime per their page.

→ opencode.ai/go

Mistral (Le Chat Pro + Vibe) — EU preference

Pick it when EU hosting, open weights, or data-handling terms matter more than raw frontier position.

What ~$15 gets you: Le Chat Pro includes Mistral Vibe (terminal coding agent). Trails closed frontier models on the hardest multi-step agent work.

→ chat.mistral.ai

Kimi — Chinese open-weight-friendly route

Pick it when you want Kimi models and a Code CLI path around the ~$19 membership band. Allowances are often unpublished — treat as “learn by using”.

→ kimi.com

GLM Coding Plan (Z.ai) — harness + cheaper model

Pick it when you want to keep a Claude Code-style harness pointed at a cheaper endpoint. Lite sits near $18/mo; Pro jumps well above a $20 budget.

→ z.ai

Devin Pro / Devin Desktop — delegate tickets

Pick it when you want to hand off whole tasks to a cloud agent and review PRs. Quotas are often unpublished; extra usage bills at API rates. Windsurf’s lineage lives here as Devin Desktop.

→ devin.ai

Muse Code — Meta CLI with a ~$15 plan

Pick it when you want Meta’s terminal agent on a flat monthly plan instead of (or before) pure API billing.

What you get: Muse Code documents subscription plans (Everyday / High / Power — commonly cited around $5 / $15 / $50) for the CLI-connected credential, plus Meta Model API pay-as-you-go for other keys. High (~$15) is the one that sits in this $20-band guide.
Watch out for: plan benefits can vary by region; confirm prices in onboarding. Extra API keys stay token-billed. Longer product notes: Muse Code vs Claude Code and Cursor.

For most vibe coders with one $20 slot, I still start with Claude Code, Codex, or Cursor unless Meta’s stack is the point.

→ dev.meta.ai/docs/muse-code

Free first, then pay

Antigravity free tier — agent product with weekly limits.
Kiro bonus / free credits — burn them on your hardest task.
Copilot Free and vendor free chat tiers — compare feel.
If ~$10 is enough to start, compare Cursor Plus ($8), Copilot Pro (editor), and OpenCode Go (agent models).
Then put the remaining budget on the ~$20 tool whose limits you actually hit — or stretch to Cursor Premium ($30) if the IDE is the job.

API-key map for free starts: Free AI API keys for vibe coding (2026).

FAQ

What is the best AI coding tool for $20 a month?

There is no single best. Claude Pro (Claude Code), ChatGPT Plus (Codex), Google AI Pro (Antigravity), and Kiro Pro all cost about $20. Cursor’s serious new-buyer plan is Premium at $30 (Plus is $8; legacy Pro $20 still exists for some accounts). Use the interactive chooser on the site, then confirm limits on the vendor page.

Cursor vs Claude Code — which should I pay for first?

Pick Cursor Premium ($30 for new buyers) if you want one AI IDE with a shared usage pool and cloud agents. Pick Claude Pro ($20) if your work is terminal-first multi-file agent sessions on Anthropic models. With a strict $20 budget, Claude/Codex/Antigravity/Kiro fit cleaner than Premium; Cursor Plus ($8) is a light Grok-oriented step. Start with the surface you live in all day.

Codex vs Claude Code — which is better on Plus or Pro?

Both sit in the $20 band and use five-hour windows plus weekly caps. Prefer Codex if you already like OpenAI models and want published limit ranges plus GitHub or Slack extras. Prefer Claude Code if Anthropic models and long multi-file refactors are your default. Neither replaces an IDE by itself.

Is Cursor unlimited on the new plans?

No. On Cursor’s new ladder (Plus $8 / Premium $30 / Super $100 / Ultra $200), included usage is a weekly or monthly pool — not unlimited agent runs. Third-party models are often billed separately at API rates. Legacy Pro ($20) for existing accounts also uses included pools plus optional on-demand. Tab completions are typically uncapped.

Does ChatGPT Plus include Codex?

Yes. Codex is included on ChatGPT Plus across web, CLI, IDE extension, and iOS, with five-hour windows and weekly caps. See the one-month Codex notes linked in this post for real-session cliffs.

Does Claude Pro include Claude Code?

Yes. Claude Code is included on paid Claude plans and shares the same five-hour and weekly limits as Claude chat. A long chat day shortens your coding day.

Is Google Antigravity free?

Antigravity has a free individual tier with basic weekly limits. Higher agent capacity comes with Google AI Pro at about $19.99 per month. Gemini 4 Argon works on paid, non-trial AI Pro in the Gemini app, AI Mode, and the Antigravity app and CLI (not the Antigravity IDE itself). AI Ultra gets 5x or 20x the Pro usage limits, and Argon in AI Studio with your plan is still Ultra-only. Start free if your stack is Google, then upgrade when the weekly cap blocks real work.

Is Kiro a good $20 pick if my stack is AWS?

Yes if you build on AWS and like writing specs and tasks before code. Kiro has a free credit path to test, then Pro at about $20 per month with a monthly credit pool. Start on free credits for your hardest task before you pay.

OpenCode Go vs GitHub Copilot Pro — which $10 plan should I buy?

Copilot Pro is the editor plugin path inside VS Code, JetBrains, and GitHub. OpenCode Go is a $10 model-access subscription for curated low-cost coding models you can use with OpenCode or other agents, with published five-hour and monthly limit tables. Pick Copilot if the IDE plugin is the job; pick Go if you want cheap agent model capacity. Go Plus is $40 if limits bite.

Originally published on Vibe Coder's Life. Run the interactive chooser there, or subscribe for the next issue.

Sources: Vendor pricing pages for Cursor, Anthropic, OpenAI, Google AI, Kiro, GitHub Copilot, OpenCode Go, Mistral, Kimi, Z.ai, Devin, Meta Muse Code (checked mid-Oct 2026) · prior VCL posts linked above.

Disclosure: No paid placement by the AI coding vendors compared in this post. Product links are plain catalog URLs. Drafting/editing used AI assistance.

My vibe coding stack in 2026: 12 tools, free starts, and paid upgrades

Viktor Kondas — Wed, 21 Oct 2026 11:00:00 +0000

Cursor, GitHub, Vercel, Neon, Resend, Stripe, Apify, Revolut, Claude, ChatGPT, Google Gemini, GoDaddy — the exact stack behind vibecoderslife.com, with honest takes, three dollar budget tiers, and the tools I actually keep.

Free vibe coding stack — **epilogue* (my personal stack).*

Parts 1–8 of Free vibe coding stack mapped the options — keys, courses, hosting, database, auth, email, monitoring, payments. This is the epilogue: what I actually open when I start a project, what you can start free, and which paid upgrade is worth it later.

Everything below is something I run on vibecoderslife.com right now, or use daily while building side products. I've skipped tools I tried and dropped. Prices checked 9 Oct 2026.

Stack in one screen

Role	My pick
Build	Cursor
Models	Claude + ChatGPT + Gemini
Code	GitHub
Deploy	Vercel
Database	Neon
Email	Resend
Payments	Stripe
Data / Actors	Apify
Banking	Revolut
Domain	GoDaddy

Auth when a product needs login: Clerk (Part 5). This marketing site does not use auth. For monitoring after you ship, see Part 7.

Here's why I chose each one.

Pick a budget

Dollar list prices; your card currency may differ.

$0 — free route

Cursor free · GitHub Free · Vercel Hobby · Neon free · Resend free · Stripe (pay when you sell) · Apify free to explore · Claude / ChatGPT / Gemini free · Revolut Standard · a domain when you need one.

Optional: saymd (free MIT CLI) and AllStackd (start on trial; paid when SES is worth it).

$20–50 — what I'd choose

Keep the free path. Pick one $20 tool first:

Cursor Pro — $20/mo (the upgrade I'd keep first)
ChatGPT Plus (Codex) — $20/mo
Claude Pro (Claude Code) — $20/mo
Google AI Pro (Gemini Antigravity) — $20/mo

When you ship for real, add Vercel Pro ($20) and Apify Creator ($6 for the first 6 months + $500 usage credits) · apify.com.

Optional: saymd / AllStackd — start free or on trial, upgrade later.

$100+

Path A: one Max / high tier — Claude Max or ChatGPT Pro/Max (~$100+/mo).

Path B: three $20 tools (Cursor Pro + Claude Pro (Claude Code) + ChatGPT Plus (Codex)) plus Vercel Pro, Apify Creator when Actors are the product, and Resend Pro / Neon Launch when free tiers are tight.

1. Cursor

What it is: An AI-native code editor built on VS Code. You describe what you want; the agent edits the repo.

Series: Part 1 · Free AI API keys · Part 2 · Courses.

Why I use it: It replaced the "open ChatGPT in a tab, paste code, copy back" loop. The agent understands the whole codebase, not just the snippet in your clipboard. I write almost no boilerplate by hand anymore.

Free start: Cursor's free tier is enough to test the workflow and build small things.

First paid tier: Cursor Pro at $20/month. This is the one upgrade I'd keep first if I had to cut everything else.

Honest take: The context window is not infinite. Large files sometimes confuse it. But for shipping marketing sites, API routes, and email templates — it's the single biggest productivity lever I've found.

→ cursor.com

2. GitHub

What it is: Version control, collaboration, and the CI trigger that deploys this site.

Series: Part 3 · Free hosting.

Why I use it: Push to main → Vercel deploys. That's the whole pipeline. I also use GitHub as the public home for my open-source work.

Free start: GitHub Free is enough for solo builders and public repos.

Honest take: If you're vibe coding without Git, one bad agent edit can nuke hours of work. Commit often.

→ github.com/kondasviktor

3. Vercel

What it is: Hosting for static sites and serverless functions. This site is plain HTML + Node api/ routes — no Next.js.

Series: Part 3 · Free hosting.

Why I use it: Zero-config deploy from GitHub. Free Hobby tier covers everything vibecoderslife.com needs. Serverless functions handle subscribe, contact, email delivery, Stripe webhook — all without managing a server.

Free start: Vercel Hobby is enough for landing pages, early APIs, and side projects.

First paid tier: Vercel Pro at $20/month when you want stronger team workflows, more scale headroom, and fewer platform limits. Hobby is also non-commercial — live charges mean Pro (Part 3).

Honest take: The Hobby tier has function count limits. I've hit them once on the download API, which I solved with a rewrite rule. For a side project with normal traffic, you won't hit any wall.

→ vercel.com

4. Neon

What it is: Serverless Postgres. Scale to zero when idle; wake on first query.

Series: Part 4 · Free database and storage.

Why I use it: No Docker, no self-managed DB, no Supabase overhead. I store subscribers, posts, and digital asset metadata in Neon. The free tier is generous — I haven't paid a cent yet.

Free start: Neon free tier is enough for a marketing site, subscriber list, and small Postgres-backed tools.

First paid tier: Launch at $15/month — that's the intermittent-load plan around 1 GB, which is the first paid step most side projects actually hit.

Honest take: Cold starts are real but unnoticeable on a marketing site. Queries run in milliseconds once awake. If you're used to SQLite locally and Postgres in prod, Neon is the smoothest bridge I've found.

→ neon.tech

5. Resend

What it is: Transactional and newsletter email. news@vibecoderslife.com runs on Resend.

Series: Part 6 · Free email and SMS.

Why I use it: Clean API, excellent deliverability, Node SDK that feels natural. The newsletter digest, welcome email, contact form replies — everything goes through Resend.

Free start: Free up to 3,000 emails/month and 100 per day. More than enough to launch.

First paid tier: Resend Pro at $20/month when your newsletter or transactional volume outgrows the free cap.

Honest take: The free tier is genuinely good for an early-stage newsletter. When I hit 10k subscribers, I'll upgrade. Not a problem I have yet.

→ resend.com

6. Stripe

What it is: Checkout, payment processing, and webhook handling for paid products.

Series: Part 8 · Payments.

Why I use it: When I launch a paid product, I need one line of code to create a Checkout session and one webhook to fulfill the order. Stripe does both. No monthly fee — you pay per transaction only.

Free start: No monthly fee. You only pay transaction fees when you actually sell something.

Honest take: Stripe is overkill if you never charge anyone. The Apify Actor Playbook is live, so Stripe is already wired in. Setup took an afternoon with Cursor doing the work.

→ stripe.com

7. Apify

What it is: A cloud platform for web scraping, automation, and data extraction. I build and publish Actors (serverless scrapers) here.

Series: How I create Apify Actors · Playbook.

Why I use it: Apify handles infrastructure, scheduling, storage, and the marketplace. I write the scraper code; Apify runs it, stores results, and handles billing for buyers. It's the platform behind the products I'm building.

Free start: Enough to explore the platform and understand the Actor model.

First paid tier: Creator at $6 for the first 6 months + $500 usage credits.

Honest take: Apify is separate from this marketing site, but it's the core of what I'm building as a business. The free tier is enough to test, and Creator is the first serious upgrade I'd recommend.

→ apify.com

8. Revolut

What it is: A personal finance app for spending, saving, and international transfers.

Series: Part 8 · Payments.

Why I use it: My Stripe payouts, tool subscriptions, and ad spend all flow through one card. Currency conversion is built in, which matters when you're paying US SaaS tools in EUR.

Free start: Standard plan.

First paid tier: Paid plans if you want extra banking features like premium cards or travel perks.

Honest take: Not a dev tool, but part of the real cost of running a side business. If you're collecting revenue from anywhere in Europe, Revolut makes the money ops simple.

→ revolut.com

9. Claude (Claude Code)

What it is: Anthropic's AI assistant. I use it daily alongside Cursor. Claude Code is the agentic coding path on Pro.

Series: Part 1 · Free AI API keys.

Why I use it: Claude Opus is my go-to for hard problems — architecture decisions, writing copy, debugging logic I can't follow, and anything requiring nuance. Sonnet handles faster drafting and email copy.

Free start: Free / limited entry path.

First paid tier: Claude Pro (Claude Code) at $20/month.

Honest take: Cursor already has Claude under the hood. But the Claude.ai interface is better for long conversations, document analysis, and anything where I want a second opinion without touching the codebase. Worth the separate subscription if you're writing and coding daily.

→ claude.ai

10. ChatGPT (Codex)

What it is: OpenAI's flagship AI assistant, with Codex as the coding agent on Plus.

Series: Part 1 · Free AI API keys · One month with Codex on ChatGPT Plus.

Why I use it: Research, image generation, Codex when I want a second coding agent, and a useful second opinion when Claude and I aren't agreeing.

Free start: ChatGPT free tier.

First paid tier: ChatGPT Plus (Codex) at $20/month.

Honest take: I use Claude more for coding and ChatGPT more for research and image tasks. They're genuinely different tools — not redundant if you're shipping daily.

→ chatgpt.com

11. Google Gemini (Antigravity)

What it is: Google's AI assistant and workspace layer across Gmail, Docs, Sheets, Slides, and NotebookLM. Antigravity is the agentic coding angle on Google AI Pro.

Series: Part 1 · Free AI API keys · Free Google AI tools.

Why I use it: It's the most practical "AI already inside your workday" option if you live in Google Workspace. Slides, Docs, Gmail summaries, and NotebookLM all make it useful beyond chat alone.

Free start: Gemini has a usable free tier for basic prompting and testing.

First paid tier: Google AI Plus at $5/month, or Google AI Pro (Gemini Antigravity) at $20/month.

Honest take: I don't use Gemini as my main coding brain, but it's very strong as an everyday assistant when your files, notes, and email already live in Google.

→ gemini.google.com

12. GoDaddy

What it is: Domain registrar. vibecoderslife.com lives here.

Series: custom domains on free hosts — Part 3 · Free hosting.

Why I use it: Honestly? It was the one I knew. DNS, renewal, email forwarding — all work fine. Nothing exciting to say here.

Free start: No free tier here — this is just the domain purchase.

First paid tier: The base domain plan is enough unless you want extra add-ons.

Honest take: The domain registrar almost doesn't matter. Pick one with clear renewal pricing and 2FA. Namecheap or Cloudflare Registrar are equally good alternatives.

→ godaddy.com

Free start vs paid upgrade

Tool	Free start	First paid tier
Cursor	Free tier to test the workflow	Pro — $20/mo
GitHub	Free for solo builders and public repos	Team if you need org workflows
Vercel	Hobby	Pro — $20/mo
Neon	Free serverless Postgres	Launch — $15/mo (intermittent load, ~1 GB)
Resend	Free up to 3k emails/mo	Pro — $20/mo
Stripe	No monthly fee	Same plan, pay per transaction
Apify	Free to explore and test	Creator — $6 for first 6 months + $500 credits · apify.com
Revolut	Standard	Paid plans if you want extra banking features
Claude (Claude Code)	Free / limited entry path	Pro — $20/mo
ChatGPT (Codex)	Free	Plus — $20/mo
Google Gemini (Antigravity)	Free	Plus — $5/mo · Pro — $20/mo
GoDaddy	Domain purchase	Higher tiers / add-ons if you need them

That's the version of the stack I would recommend to most readers: start free almost everywhere, then pay only where the upgrade clearly removes friction.

What I built with this stack

saymd — Speak once, get a structured Markdown prompt file you can @ into Cursor, Claude Code, or any coding agent. Free MIT CLI with your own API key.
AllStackd — Amazon SES control plane. Monitor every account, region, and domain from one workspace. Delivery stays in your AWS.
VCL VibeBench — Run the same Fun, Dev, and Score prompts across models yourself — no blended “best model” number.

Five free repos (with why)

VCL AI Coding Prompts — 50 prompts + agent-rules files I actually use.
VCL AI Image Prompts — 117 copy-paste image prompt recipes for ChatGPT, Gemini, Grok, and Microsoft AI.
VCL Model Arena — Run models yourself rather than trusting benchmark slides.
Apify Actor Starter Kit — Start an Actor without figuring out the plumbing from scratch.
Awesome Apify Actor Ideas — Free niche list; the scored 100 live in the Playbook.

The Apify Actor Playbook

The Vibe Coder's Playbook — Build Profitable Apify Actors with AI is the paid toolkit I use to ship Apify Actors with Cursor.

It includes:

Step-by-step Actor development with Cursor + the Apify SDK
Ready-made .cursorrules and .clauderules templates tuned for Apify scraping projects
Actor monetisation strategy: niche selection, pricing, Store listing optimisation
Real examples from the Actors I've built and published

Price: €37.

→ Get the playbook

FAQ

What is the cheapest vibe coding stack that can still ship a live site?

A $0 path works for a marketing site: Cursor free, GitHub Free, Vercel Hobby, Neon free, Resend free, Stripe (pay when you sell), Apify free to explore, free Claude / ChatGPT / Gemini tiers, and Revolut Standard. Limits bite when usage grows.

Which paid upgrade should I buy first for vibe coding?

In the $20–50 band, pick one $20 tool first — Cursor Pro is the upgrade I would keep first — or ChatGPT Plus with Codex, Claude Pro with Claude Code, or Google AI Pro with Gemini Antigravity. Add Vercel Pro and Apify Creator when Hobby limits or Actors force it.

Do I need Vercel Pro on day one?

No. Vercel Hobby covers landing pages and early APIs. Upgrade to Pro at $20/month when Hobby is non-commercial for live charges or you hit function limits. Part 3 covers hosting in detail.

Cursor Pro vs Claude Code vs ChatGPT Codex — which should I use?

They overlap but are not the same. Cursor Pro is the editor agent in your repo. Claude Pro unlocks Claude Code. ChatGPT Plus unlocks Codex. On a $20–50 budget, pick one first; on $100+ you can combine three $20 tools or buy a Max plan.

Is Neon free enough for a newsletter site?

Yes for early traffic. This site stores subscribers, posts, and digital asset metadata on Neon free. Launch at $15/month is the first paid step when intermittent load or storage needs grow. See Part 4 for the database map.

Where do Apify and Revolut fit if I am not scraping yet?

Apify is for Actors and data products when scraping is the product. Revolut sits next to Stripe for payouts and multi-currency SaaS bills. Both appear in Part 8 payments context; Apify Creator adds usage credits when you are ready to ship Actors.

What free repos come with this stack?

Five public repos: VCL AI Coding Prompts, VCL AI Image Prompts, VCL Model Arena, Apify Actor Starter Kit, and Awesome Apify Actor Ideas. The scored Actor ideas list lives in the paid Playbook.

How does this epilogue relate to Free vibe coding stack Parts 1–8?

Parts 1–8 map options for keys, courses, hosting, database, auth, email, monitoring, and payments. This post is the personal epilogue: the short list I actually run on vibecoderslife.com, with dollar budgets and links back to each part.

Originally published on Vibe Coder's Life. Tuesday/Friday digest on AI coding tools — subscribe.

Sources: Free vibe coding stack Parts 1–8 · Cursor · Vercel · Neon · Resend · Stripe · Apify · Revolut · saymd · AllStackd · five GitHub repos linked above. Prices checked 9 Oct 2026.

Disclosure: Apify and Revolut links are referral / partner links. saymd and AllStackd are my own products. Cursor, Claude, ChatGPT, Gemini, Vercel, Neon, Resend, Stripe, GitHub, and GoDaddy links are plain product URLs. Drafting/editing used AI assistance.

5 Free Google AI Tools for Vibe Coding

Viktor Kondas — Mon, 19 Oct 2026 11:00:00 +0000

Setup guides, export paths, and copy-paste prompts you can try today with a free Google account — built to share, not just read.

Looking for free Google AI tools worth opening this weekend? Start with Antigravity, Stitch, Pomelli, Flow & Flow Music. One Google account unlocks an agent coding stack, a UI generator, on-brand marketing creatives, a video studio, and a music studio. The prompts below are generic on purpose — drop them in, hit generate, and see what sticks.

Google Labs — home for Flow, Stitch, Pomelli, Flow Music, and more experiments.

Quick map

#	Product	What it is for	Start
1	Antigravity	Agent platform (desktop, CLI, IDE, SDK)	antigravity.google
2	Stitch	Text / sketch → UI screens	stitch.withgoogle.com
3	Pomelli	On-brand marketing from your site URL	labs.google.com/pomelli
4	Google Flow	Stories, image→video, cinematic clips	labs.google/fx/tools/flow
5	Google Flow Music	Full songs + music videos (Lyria 3.5)	flowmusic.app

Honest caveat: Labs experiments change fast — free tiers, waitlists, and regional availability can shift. Treat this as a discovery guide, not a SLA.

1. Antigravity — four surfaces, one agent harness

Google Antigravity is Google’s agentic development platform. You are not locked into one window: Antigravity is accessible through four distinct surfaces. Depending on whether you prefer terminal speed, a visual command center, a full code editor, or programmatic automation, pick the interface that matches your style.

Official Antigravity homepage — download the desktop app or explore use cases.

The 4 ways to use Antigravity

Surface	Form factor	Best for	Key capabilities
Antigravity 2.0	Standalone desktop app (macOS, Windows, Linux)	Visual task orchestration & async workflows	Multi-project / workspace management; visual artifacts (diffs, UI screenshots, browser recordings); scheduled background tasks & voice input
Antigravity CLI	Lightweight terminal UI (`agy`)	Terminal-centric work, SSH, fast edits	Keyboard-driven TUI in shell/tmux; near-zero footprint; shared settings + conversation export to the desktop app
Antigravity IDE	Full AI-native code editor	All-in-one primary coding environment	Built on the open-source VS Code foundation; side-by-side Agent Manager + text editor; deep repo context
Antigravity SDK	Python programmatic framework	Custom agents & automated pipelines	Python library on the Antigravity harness; register tools, safety policies, spawn subagents in code

The “dual-wield” setup

If you already love VS Code or Cursor, you do not have to switch editors completely. The recommended setup is dual-wielding:

Keep your standard editor open for everyday writing and quick tweaks.
Run Antigravity 2.0 (desktop) or the CLI in parallel for autonomous multi-file refactoring, background testing, and long-running agent tasks.

Same idea if you prototype in the browser with Replit — keep Replit for quick experiments, hand heavy refactors to Antigravity.

How to set up (desktop + CLI)

Open antigravity.google → Download for your OS (Apple Silicon / Intel / Windows / Linux as offered).
Install Antigravity 2.0 or the IDE build you prefer; sign in with your free Google account.
Open a project folder (same mental model as a VS Code workspace).
Optionally install the CLI for terminal / SSH sessions:

macOS / Linux:

curl -fsSL https://antigravity.google/cli/install.sh | bash

Windows PowerShell:

irm https://antigravity.google/cli/install.ps1 | iex

Launch and authenticate (browser OAuth on first run):

agy

/config
/logout

SDK users: follow the Python harness docs on the Antigravity site once desktop/CLI auth works — same account, programmable agents.

Starter prompts (paste into 2.0, IDE, or `agy`)

Scan this repo. Summarize the stack in 5 bullets, list the top 3 risks for a solo builder, and propose a 1-hour fix list. Do not change files yet.

Add a /health JSON route if missing. Keep the change minimal, write a short curl check, and show me the diff before applying anything destructive.

Create or update AGENTS.md with project conventions: never commit .env, prefer small PRs, explain folder layout in under 40 lines.

2. Stitch — text or sketch → UI screens

Google Stitch is an AI-native UI design tool from Google Labs. Powered by Gemini models, it skips the traditional blank canvas: you describe what you want (text or voice) or upload a rough sketch / wireframe / screenshot, and Stitch returns structured, responsive screens in seconds.

Two modes in practice: Standard (fast text-to-UI) and Experimental (stronger sketch-to-UI translation). Use it for mobile shells, web dashboards, landing pages, and multi-screen first drafts (splash → signup → home → settings).

Stitch — Design with AI (official marketing asset).

What it can generate

Mobile: iOS / Android viewports, bottom nav, card stacks, onboarding, checkout, widget dashboards.
Web: SaaS dashboards, admin panels, landing pages, pricing grids, settings, data tables.
Flows: multi-screen journeys as a first draft you refine, not a final design system.

Export options (and where to take the UI next)

Platform	Method	How it works
Figma	Native “Copy to Figma”	Paste (`Cmd/Ctrl+V`) into Figma for editable layers / Auto Layout. Standard Mode.
HTML / CSS / Tailwind	Native code export	Copy modular front-end markup into VS Code, Cursor, or any IDE.
Bolt.new & Lovable	Code / spec paste	Drop Stitch HTML/Tailwind or a short `design.md` into Bolt or Lovable to turn static UI into a runnable app with state.
Google AI Studio & Firebase Studio	Google ecosystem	Carry designs / schemas into AI Studio prompts or Firebase Studio for auth, Firestore, and backend wiring.
v0	Prompt + markup	Paste the layout brief (or exported markup) into v0 when you want React components next.

Setup: open stitch.withgoogle.com, sign in with Google, paste one of the prompts below.

5 UI ideas & copy-paste prompts

Clear hierarchy + palette + layout rules beat vague “make it pretty” prompts.

Idea 1 — Dark mobile finance / crypto dashboard

Tests dense data, charts, and horizontal cards on mobile.

Create a modern mobile iOS dashboard for a crypto tracking app called 'Aura Crypto'.
Layout:
- Top Bar: Avatar icon on the left, notification bell with badge on the right.
- Hero Card: Dark slate background with a subtle neon green gradient border. Show total balance '$42,580.12' in 32px bold text, with a '+4.2%' pill badge.
- Quick Actions: Row of 4 circular icon buttons ('Send', 'Receive', 'Swap', 'Buy').
- Horizontal Scroll Section: 'Watchlist' showing mini-sparkline charts for BTC, ETH, and SOL.
- Bottom Navigation: 4 tabs ('Home', 'Markets', 'Analytics', 'Profile') with high-contrast active state.
Theme: Deep dark mode (#0D1117 background) with neon accent colors.

Idea 2 — Web SaaS analytics + data table

Tests desktop columns, filters, and dense tables.

Generate a desktop web application UI for an AI Customer Support Analytics dashboard.
Layout:
- Sidebar (Left, 240px wide): Logo at top, dark navigation links ('Overview', 'Tickets', 'CSAT Scores', 'Team Performance', 'Settings').
- Main Header: Search bar with hotkey indicator '⌘K', date range picker dropdown ('Last 30 Days'), and export CSV button.
- Top Metric Grid: 4 stat cards ('Total Tickets', 'Avg Resolution Time', 'CSAT Score 94%', 'AI Deflection Rate 68%') with small trend arrows.
- Main Section: Split 70/30. Left side shows a line chart area for 'Ticket Volume over Time'. Right side shows a 'Live Activity Feed' list.
- Bottom Table: Paginated user tickets table with status tags ('Open' in yellow, 'Resolved' in green, 'Escalated' in red).
Theme: Light clean SaaS theme, neutral cool gray canvas with indigo accent buttons.

Idea 3 — Mobile health onboarding step

Tests progress UI and selectable cards.

Design a mobile UI screen for Step 2 of a personalized fitness onboarding app.
Layout:
- Top Bar: Back arrow on left, progress bar showing 'Step 2 of 5' (40% filled).
- Headline: Large 28px bold title 'What is your primary fitness goal?' with a friendly subtitle 'We'll adapt your daily workout routines accordingly.'
- Options Stack: 4 vertical clickable cards with rounded corners:
  1. 'Build Muscle & Strength' (Include a dumbbell icon)
  2. 'Lose Weight & Burn Fat' (Include a flame icon)
  3. 'Improve Endurance' (Include a running shoe icon)
  4. 'General Health & Flexibility' (Include a yoga icon)
- Make card #1 look 'Selected' with a prominent green border and a checkmark.
- Bottom Bar: Fixed 'Continue' primary button spanning full width with subtle drop shadow.
Theme: Minimalist white background, bold black typography, emerald green primary accents.

Idea 4 — E-commerce product detail (web)

Tests gallery grids and sticky CTAs.

Create a desktop web UI layout for a premium wireless headphone product detail page.
Layout:
- Top Nav: Sticky transparent navbar with brand logo, shop links, search, and shopping bag icon with item count '2'.
- Split View (50/50):
  - Left Column: Large main product image showcase with 4 smaller thumbnail preview squares below it.
  - Right Column: Product title 'Apex Wireless ANC Headphones', star rating (4.9/5 - 320 reviews), price '$299.00', color swatch selectors (Space Gray, Silver, Midnight Blue).
- Product Details: Collapsible accordion menus for 'Technical Specs', 'What's in the Box', and 'Shipping & Returns'.
- Primary Action: Large full-width 'Add to Cart' button next to a quantity selector (+ / -).
Theme: Minimalist luxury aesthetic, off-white background, crisp dark typography, soft shadows.

Idea 5 — Telehealth video call (mobile)

Tests overlays, PiP, and floating controls.

Design a mobile UI screen for an active telehealth video consultation call.
Layout:
- Background: Full-screen remote doctor video view.
- Top Floating Bar: Doctor name 'Dr. Sarah Chen, MD - Cardiology', call timer '12:44', and end call red button.
- Picture-in-Picture: Small rounded thumbnail in top-right showing patient's own camera view.
- Floating Bottom Overlay Bar: Semi-transparent blurred glassmorphism bar with 4 action buttons: 'Mute Mic', 'Turn Off Camera', 'Open Chat', 'Prescriptions'.
- Slide-up Drawer (Half Open): Shows recent shared doctor notes: 'Recommended Dosage: 50mg daily' with a download PDF button.
Theme: Clean clinical aesthetic with semi-transparent frosted glass UI elements.

3. Pomelli — paste your site URL, get a brand kit + campaigns

Pomelli (Google Labs + DeepMind) is built for small and mid-size businesses that need on-brand social creatives without a full design team. The core loop is simple: you give it your website URL, it crawls the brand, then builds a mood board and analysis — colors, tone, typography cues, and visual direction — before you generate campaigns.

Pomelli — official Google Blog header art.

How it actually works

Open labs.google.com/pomelli/about → Let’s get started (sign in with Google; early beta rolled out in English for US, Canada, Australia, New Zealand first).
Enter your site URL. Pomelli analyzes the brand and produces a mood board / style summary (palette, voice, visual motifs).
Create a campaign: pick output ratio (square, story, landscape, etc.), type what to generate, and tweak with simple instructions — change the background, swap a product photo, lean into a holiday, etc.
Use presets and event angles: product mockups, Mom’s Day, Father’s Day, seasonal sales, launch teasers, and other campaign templates the product surfaces.

Official Pomelli demo still from the Google Blog (brand creative workflow).

Official Pomelli demo still — campaign / style variations.

From Pomelli assets → social scheduling

Once you like the images, schedule them. We use OmniSocials for that layer: one workspace to queue posts across Facebook, Instagram, Pinterest, LinkedIn, YouTube, TikTok, X, Threads, Bluesky, and more — plan captions, approve, and publish on a calendar. Sign up via that link if you want to try the same stack after Pomelli exports.

Brand site: https://your-business-site.example
Campaign: Summer launch — 3 square posts + 2 story frames.
Tone: friendly, clear, no hype.
Instructions: Keep brand colors from the site analysis. Swap background to soft daylight. Add 'Free shipping this week' as secondary text on one square.

4. Google Flow — AI creative studio for stories & video

Google Flow is Google’s AI creative studio for cinematic scenes, storyboards, image-to-video, and narrative clips. Multimodal models (including Gemini-class and Veo video engines) let you combine text, stills, and camera controls into continuous shots.

Core moves: image→video with camera physics (dolly, pan, orbit); chain shots for consistent lighting / character; agentic tool pipelines when you want repeatable creative utilities.

Google Flow landing page on Labs.

Open labs.google/fx/tools/flow.
Sign in with Google and accept Labs terms.
Start from a still or a short story brief — iterate instead of one perfect prompt.

Image-to-video prompts to try

Use case A — cinematic character motion

Subject: The character in the input image.
Motion: Slow orbit camera shot moving in a tight circular arc around the subject. The subject turns their head slowly toward the camera and settles with a subtle smile blooming.
Atmosphere: Soft golden-hour sunlight shifts naturally across their face, fabric of the coat sways gently in a light breeze.
Camera: Eye-level, shallow depth of field, blurred background bokeh.

Use case B — environment & lighting transition

Subject: The landscape from the input image.
Camera Motion: Smooth dolly forward tracking shot pushing gradually deeper into the scene over 5 seconds.
Environment Motion: Thin mist drifts slowly across the water surface, warm sunlight refracts through rising atmospheric fog, and tree leaves rustle with natural weight.
Mood: Inspiring, serene, atmospheric.

5. Google Flow Music — songs + music videos (Lyria 3.5)

Google Flow Music is the music studio companion: chat with an AI producer, generate ~30-second to multi-minute tracks from text, lyrics, images, or audio refs, and direct music videos. Free to start with a Google account on the public landing.

As of 29 July 2026, Flow Music is rolling out Lyria 3.5 — Google DeepMind / Labs’ newest music generation model. Google’s pitch is richer tracks with more creative control, not just “another song button.”

What’s new in Lyria 3.5 (per Google’s launch post):

Improved musicality — richer, more complex melodic structures that sound more natural.
Enhanced lyrics — higher-quality lyrics with better prompt adherence and structural awareness (verse / chorus shape actually sticks).
Improved vocals — more expression and emotion, more realistic / nuanced delivery, plus better pronunciation.
Creative control — easier control of tempo and duration on outputs.

Core moves in the studio: conversational arrangement tweaks (tempo, instruments, vocals); structural tags like [Verse 1] / [Chorus]; auto cover art generation. With Lyria 3.5, those tempo / length knobs and lyric structure tags are the levers worth leaning on first.

Google Flow Music — flowmusic.app (Lyria 3.5 in the producer stack).

Go to flowmusic.app.
Sign up / sign in with Google.
Use a starter or chat with Producer — keep genre, BPM, duration, and structure explicit so Lyria 3.5’s control improvements matter.

Music prompts to try

Use case C — full song with structure tags

Style: Synthwave, Cyberpunk, 80s Retrowave
Tempo: 115 BPM
Instrumentation: Analog poly-synths, heavy sidechained bassline, gated reverb snare, soaring lead guitar.
Mood: Energetic, nostalgic, late-night drive.

[Intro]
(Atmospheric synth pads building with slow sidechain pulsing)

[Verse 1]
Neon lights reflecting on the rainy street below,
Chasing digital shadows where the city lights glow.
Engine revs in silence, counting down the time.

[Chorus]
We run through the static, fading in the light,
Speeding through the grid into the endless night!

Use case D — ambient / lo-fi instrumental

Style: Lofi Chillhop, Jazztronica
Tempo: 78 BPM
Instrumentation: Dusty Rhodes electric piano, vinyl crackle, soft brushed hip-hop drums, warm upright bass line.
Mood: Cozy, focused, rainy afternoon vibe.
Instruction: Instrumental only. No vocals. Smooth intro fade-in, organic loop structure.

Combined multimodal workflow (image → video → soundtrack)

[Step 1: Visual Prompt]
A futuristic astronaut standing on an alien desert dune looking at a glowing blue galaxy on the horizon. Wide cinematic 35mm shot.

[Step 2: Video Motion Prompt]
Slow crane up camera shot rising vertically to reveal the vast expanse of the desert. Dust particles float in slow-motion, galaxy shifts with subtle aurora-like energy movement.

[Step 3: Music Prompt]
Cinematic ambient sci-fi score. Deep sub-bass drones, shimmering orchestral strings, high-pitched pitch-bent synths. Building intensity, inspiring sense of awe and wonder.

Run step 1–2 in Flow, step 3 in Flow Music, then pair the audio under the clip — that three-step loop is the viral share people actually try.

How the five fit together

Antigravity — ship and refactor code (desktop / CLI / IDE / SDK).
Stitch — invent UI, then export to Figma, code, Bolt, Lovable, or v0.
Pomelli — brand analysis from your URL → campaign creatives → schedule with OmniSocials.
Flow — turn stills into motion.
Flow Music (Lyria 3.5) — score the story.

What do you think?

Which prompt did you try first — and what broke or surprised you? Browse more builders’ tools on our Tools page, or subscribe for the next toolkit.

Originally published on Vibe Coder's Life.

Sources: Google Labs; Google Antigravity + antigravity-cli; Stitch; Pomelli; Google Flow; Google Flow Music; Introducing Lyria 3.5 in Google Flow Music (Google Blog, 29 Jul 2026).

Disclosure: Screenshots and marketing images from official Google pages for editorial illustration. No paid placement by Google. Replit and OmniSocials links above are referral/affiliate. Cursor and v0 link to their homepages. Product availability and free tiers can change. Drafting/editing used AI assistance.

Payments for vibe coding (2026)

Viktor Kondas — Fri, 16 Oct 2026 11:00:00 +0000

How to charge for a vibe-coded project — Stripe Checkout as the start-here (transaction fees apply), Polar / Lemon Squeezy / Paddle as Merchant of Record, PayPal for wallets, BLIK and SimplePay for CEE, plus Wise and Revolut for near-zero banking FX.

Free vibe coding stack — **Part 8* (payments).*

This is the last chapter of Free vibe coding stack. The title is Payments on purpose: Stripe (and every other processor) still takes a percentage + fixed fee on each successful sale. What you avoid is a monthly platform subscription on standard pricing — not 0% take. This site’s Playbook already uses Stripe Checkout plus /api/stripe-webhook.

Part 3 reminder: Vercel Hobby is non-commercial. Processing payment is the Pro trigger. Test mode is fine on Hobby; live charges are not.

Start-here: Checkout + webhook → Stripe (Hungary: 1.5% + 85 Ft standard EEA cards; no monthly fee). Merchant of record (they file VAT) → Polar / Paddle / Lemon Squeezy / Creem. Poland → enable BLIK on Stripe. Hungary local acquiring → SimplePay (OTP) when Stripe is not the shopper’s expectation. Banking FX / multi-currency after payout → Wise or Revolut.

Processor / MoR map

Provider	Best for	Take / fee	Monthly?	Cliff
Stripe	Checkout, subscriptions, webhooks	HU: 1.5% + 85 Ft EEA standard	No	You are merchant of record; Tax +0.5% where you collect
Polar	OSS / SaaS MoR, public rate card	Starter 5% + 50¢	Starter $0; Pro $20	New orgs (after 27 May 2026) start at 5%+50¢, not 4%
Lemon Squeezy	Digital MoR, license keys	5% + 50¢	No	Still a MoR product (Link, LLC / Stripe-owned brand)
Paddle	MoR + buyer support	5% + 50¢	No	Custom quote under $10 SKUs
PayPal	Consumer wallet reach	HU: 3.40% + fixed (other commercial)	No	Extra button, not a SaaS webhook replacement
Wise	Multi-currency account + FX after payout	Transparent FX (not a Checkout % cut)	No*	Banking rail — does not replace Stripe webhooks
Revolut	Business / personal FX and cards	Plan-dependent; FX often near mid-market	Plan*	Same: hold/move money, not MoR Checkout

PSP/MoR and banking fees re-checked 28 Sep 2026. US blogs quoting 2.9% + 30¢ are not Hungary. Polar Early Member (4% + 40¢) is only for orgs created before 27 May 2026. *Wise/Revolut “monthly” depends on plan — the point is FX, not a card-processing take rate.

CEE map

Method	Best for	How	Cliff
BLIK	Polish instant pay	Enable in Stripe Dashboard (Checkout / Elements)	Presentment PLN; 6-digit code, ~2 min validity
SimplePay	Hungarian OTP acquiring	Merchant contract with OTP Mobil	Quoted per merchant — no public Stripe-style list

Hungarian Stripe accounts can accept BLIK. This site is Stripe + Számlázz.hu for NAV — not SimplePay. Do not pretend otherwise.

Wire it from Cursor

Wire Stripe Checkout using STRIPE_SECRET_KEY and STRIPE_WEBHOOK_SECRET from .env.
Walk me through: (1) test-mode keys and a Price ID,
(2) putting secrets in .env and .gitignore,
(3) the same names on the host dashboard,
(4) a Checkout Session + webhook that fulfills only after payment_status is paid.
Do not put STRIPE_SECRET_KEY in client code.
Do not go live on Vercel Hobby (Part 3 commercial rule).
If we need BLIK, enable it in Stripe — do not add a second Polish PSP.

Stripe (start-here)

Stripe Hungary: no setup fee, no monthly fee. Published card rates include 1.5% + 85 Ft standard EEA, 2.8% + 85 Ft premium EEA, 2.5% + 85 Ft UK, 3.15% + 85 Ft international, plus 2% if currency conversion applies. BLIK is a Dashboard toggle for Checkout and Elements (not Express Checkout). Stripe Tax is 0.5% on transactions where you collect. You file VAT unless you add Tax or a MoR.

Activate the account → test keys → STRIPE_SECRET_KEY, publishable key, webhook secret.
Create a Product/Price. Checkout Session on the server. Fulfill in the webhook, not in the success URL alone.
Enable BLIK if you sell to Poland. Local payment methods: Dashboard → Settings → Payment methods.

Polar

Polar is Merchant of Record with a public rate card. Starter (orgs on/after 27 May 2026): 5% + 50¢, $0 monthly. Pro $20/mo then 3.8% + 40¢. Growth/Scale lower the cut further. International cards +1.5%. Polar is “lowest SaaS fees” after you pay the monthly, or if you are still on Early Member (4% + 40¢). New indie orgs on Starter are in the same 5%+50¢ band as Paddle/Lemon Squeezy.

Lemon Squeezy

Lemon Squeezy: 5% + 50¢, no monthly ecommerce fee, MoR (VAT/sales tax). Footer in 2026: Sold through Link, LLC f/k/a Lemon Squeezy LLC — Stripe-owned, still a separate checkout. Fine for license keys and digital downloads. Same MoR trade as Paddle: higher take, less tax admin.

Paddle

Paddle: 5% + 50¢ per Checkout transaction, tax compliance and buyer billing support included. Custom pricing if the SKU is under $10. Use it when you want MoR and already like Paddle’s SaaS catalog rules. Newer EU MoR option with agent/MCP tooling: Creem (lists 3.9% + $0.40 — verify on their pricing page).

PayPal

Add PayPal when buyers insist. Do not replace Stripe webhooks with “IPN and hope.” Official Hungary merchant fees (table dated 30 Apr 2026) list 3.40% + fixed fee for “all other commercial transactions,” plus extras for international. Quote that PDF, not a US blog.

BLIK

Polish bank-app pay: customer types a 6-digit code, approves on the phone. Prefer Stripe’s BLIK if the Stripe account’s country list includes you (Hungary does). Presentment currency is PLN. Recurring BLIK exists as a Stripe preview — check docs before you promise subscriptions.

SimplePay

SimplePay is OTP Mobil’s Hungarian gateway. Shoppers know the OTP redirect. Fees are a merchant quote, not a public 1.5%+85 Ft page. Third-party roundups often cite roughly 0.9–1.6% for HU cards — treat that as unofficial. Choose SimplePay when you need local acquiring, SZÉP, or Stripe onboarding is the wall. Different dashboard, payout, and VAT ops. This site does not use SimplePay.

Wise and Revolut (near-zero banking FX)

Card Checkout and Merchant of Record solve taking money. Wise and Revolut solve holding and moving it afterward — contractor payouts, EUR↔HUF, multi-currency balances — often far cheaper than legacy bank FX. They are not a Stripe webhook replacement.

Wise — multi-currency account + transparent transfer fees (referral).
Revolut — spending, saving, and international transfers with competitive FX (referral).

Pattern that works for this site’s stack: Stripe (or a MoR) collects → payout to your bank / Wise / Revolut → pay contractors and tools from there.

Fees vs monthly (the honest cliffs)

Stripe — always a % + fixed on successful charges; monthly is $0 until you add Tax, Billing, Radar extras.
MoR — you pay ~4–5% so they file VAT (Creem lists 3.9% + $0.40). Polar Pro is how the Polar % drops.
Host — live charges on Vercel → Pro (Part 3).
SimplePay — contract + integration time, not a stripe npm install.
Wise / Revolut — FX and balances; still not “free money,” but usually cheaper than incumbent bank FX for EU builders.

Practical tips

Webhook is the source of truth — never fulfill from the thank-you page alone.
Test clocks / test cards — live mode last.
One primary PSP — Stripe or a MoR. PayPal as an extra method is fine.
NAV / invoices — Stripe’s receipt is not a Hungarian invoice. This site uses Számlázz.hu beside Stripe.
Price in the fee — if Stripe takes ~1.5% + 85 Ft, your €37 digital product is not €37 net. Model net before you launch.

That’s the Free series

Parts 1–8: keys, courses, hosting, database, auth, email, monitoring, payments. Epilogue: My vibe coding stack in 2026.

Browse tools we actually use on vibecoderslife.com/tools. Stripe is a plain product URL there; Wise and Revolut are marked referrals.

FAQ

Does Stripe charge a monthly fee?

No on standard pricing. Stripe has no setup fee and no monthly platform fee. You pay a percentage + fixed fee per successful charge (Hungary: 1.5% + 85 Ft for standard EEA cards). Optional add-ons like Stripe Tax (+0.5%) or Billing extras can raise the bill — but the baseline is pay-when-you-get-paid.

Are payments for vibe coding really free?

No. This guide’s “no monthly fee” framing is not 0% take. Every successful card charge still pays a cut. Merchant-of-record tools (Polar, Paddle, Lemon Squeezy, Creem) take a larger cut because they file VAT for you. Model net revenue before you price a €37 digital product.

What is a merchant of record vs a payment gateway?

A payment gateway like Stripe moves money while you stay the legal seller (tax registration and invoices stay yours — this site pairs Stripe with Számlázz.hu for NAV). A merchant of record (Paddle, Polar, Lemon Squeezy, Creem) is the legal seller: they collect and file VAT/sales tax, usually around 5% + 50¢ (Creem lists 3.9% + $0.40).

Does Stripe Tax file my VAT returns?

No. Stripe Tax calculates and collects the right rate and helps track nexus, but you (or your accountant) still register, file, and remit. If you want filing handled for you, use a merchant of record instead of (or on top of) Stripe Tax.

Stripe vs Paddle vs Lemon Squeezy — which should I use?

Stripe if you want Checkout, subscriptions, and webhooks and will handle tax (this site). Paddle or Lemon Squeezy if you want them to be merchant of record for global digital sales. Polar sits in the same MoR band for new orgs (~5% + 50¢ on Starter). Lemon Squeezy is Stripe-owned but still a separate MoR checkout (Link, LLC).

Is the 5% merchant of record fee worth it for a side project?

Usually yes for a solo vibe coder selling globally: the 2–3% premium vs Stripe often costs less than DIY VAT registrations and filings. Switch to Stripe later if revenue and a finance workflow justify owning tax yourself. Migration means new checkouts and often re-collecting cards — plan it.

How do I add Stripe Checkout to a vibe-coded project?

Create a Stripe account, put test keys in .env, create a Checkout Session on the server, and fulfill only from a signed webhook after payment_status is paid. Paste the Cursor prompt above.

Do I need SimplePay if I already have Stripe?

Not for a global digital product. Yes if Hungarian shoppers bounce without OTP, or you need SZÉP / local acquiring Stripe does not cover. Fees are a merchant quote — there is no public Stripe-style list.

Can I accept BLIK without a Polish company?

Stripe documents BLIK for accounts in many countries including Hungary. Enable the method in the Dashboard; presentment is PLN. Confirm in your own Dashboard — do not invent a second Polish PSP.

Where do Wise and Revolut fit if Stripe already takes a card fee?

Wise and Revolut are banking / multi-currency tools for holding and moving money with transparent FX — not a Checkout replacement. Use them after Stripe or a MoR pays you out, or for contractor payouts. They can run near-zero for many EUR/HUF transfers compared with legacy bank FX.

Originally published on Vibe Coder's Life. Tuesday/Friday digest on AI coding tools — subscribe.

Sources: Stripe Hungary pricing / BLIK; Polar; Lemon Squeezy; Paddle; Creem; PayPal HU fees; SimplePay; Wise; Revolut. Numbers re-checked 28 Sep 2026 — verify in-product.

Disclosure: Wise and Revolut links are referrals (same as on our Tools page). Stripe, Polar, Lemon Squeezy, Paddle, Creem, PayPal, and SimplePay are plain product URLs. No paid placement. This site uses Stripe Checkout in production (Playbook). Drafting/editing used AI assistance.

Free monitoring and analytics for vibe coding (2026)

Viktor Kondas — Wed, 14 Oct 2026 11:00:00 +0000

How to see traffic and errors on a vibe-coded project — Google Search Console and GA4 as the start-here, then PostHog, Sentry, Better Stack, Axiom, Plausible/Umami, and UptimeRobot.

Free vibe coding stack — **Part 7* (monitoring & analytics).*

Parts 3–6 gave you a URL, a database, login, and mail. Without measurement, you are guessing. Part 7 is monitoring and analytics. This site already runs Google Analytics 4 and Google Search Console — that is the start-here, not PostHog.

Traffic/SEO is one table. Product, errors, logs, and uptime are another. Do not run GA + Plausible + PostHog on the same page “for completeness.”

Start-here: Will Google show the URL → Search Console (free). Who visited → GA4 (free standard). Product/errors → PostHog or Sentry. Is it up → UptimeRobot or Better Stack.

Traffic / SEO map

Tool	Best for	Free quota	Card?	Cliff
Search Console	Index, queries, sitemap	Free	Google account	Not session analytics
GA4	Traffic on this site’s stack	Free standard property	Google account	Cookie consent in the EU; GA360 is paid
Plausible / Umami	Privacy-first alt	Self-host OSS	No (self-host)	Plausible Cloud is paid; pick one stack vs GA

Checked 21 Sep 2026. Microsoft Clarity (free heatmaps) is a sibling script to GA — not a third analytics religion. This is not a GTM tutorial; gtag on the page is enough to start.

Product / errors / uptime map

Tool	Best for	Free quota	Card?	Cliff
PostHog	Product + errors in one	1M events; 100k exceptions	No	1 project on Free; then pay-as-you-go
Sentry	Error tracking	5,000 errors; 1 user	No	Team from $26/mo; 30-day lookback on Developer
Better Stack	Errors + uptime + status	100k exceptions; 10 monitors	No	Free copy is personal projects; 3-day log retention
Axiom	Logs	500 GB ingest / mo (Personal)	No	10 GB-hours query; 25 GB storage; Cloud $25/mo
UptimeRobot	Ping the URL	50 monitors; 5-min interval	No	Hobby / non-profit positioning; Solo for faster checks

Uptime pings ≠ error tracking. A 200 OK can still be a broken checkout.

Wire it from Cursor

Add Google Analytics 4 using measurement ID G-XXXXXXXX from .env (or a documented constant).
Walk me through: (1) Analytics account → GA4 property → Web data stream,
(2) a single gtag snippet in the HTML head — no extra tag managers,
(3) Search Console Domain property → DNS TXT verify → submit sitemap.xml,
(4) URL Inspection on the homepage: Test live URL, then Request indexing if needed,
(5) do not add Plausible or PostHog in this change unless I ask.
Never commit a secret that is not meant to be public; GA measurement IDs are public-by-design.

Google Search Console — setup (start-here, SEO)

Search Console is free. It answers “will Google show this URL?” — queries, index coverage, sitemaps — not button clicks. Pair it with GA4. Steps below follow Google’s own help: add a property, verify ownership, submit a sitemap, URL Inspection.

1. Open Search Console and add a property

Sign in at search.google.com/search-console with the Google account that should own the site.
Click Add property.
Prefer a Domain property (example.com, no https://, no www). That covers http/https and all subdomains in one place. Domain properties verify with DNS only.
If you only control a single host path (or cannot edit DNS yet), use a URL-prefix property (https://www.example.com/) and pick HTML file, HTML tag, GA, or GTM verification — but Domain + DNS is the cleaner long-term choice for a vibe-coded site on your own domain.

2. Verify ownership (Domain = DNS TXT)

In the Domain flow, Google shows a TXT (or CNAME) record.
Open your DNS host (Vercel Domains, Cloudflare, Namecheap, …) and add that record on the apex.
Wait until the record resolves — Google notes this can take up to a few days for some providers. Then click Verify in Search Console.
Leave the TXT record in place. Removing it later can drop verification.

Stuck? Re-check the exact hostname and that you pasted the full token. Google’s verification guide lists common DNS mistakes.

3. Publish a sitemap Google can fetch

Ship a public sitemap on the live host — usually https://example.com/sitemap.xml (or a sitemap index). It must not require login and must not be blocked by robots.txt.
In Search Console, open Sitemaps.
Paste the exact sitemap URL (redirects are not followed in that field) and click Submit.
Optional: confirm fetch with URL Inspection on the sitemap URL — you want crawl allowed and page fetch successful.

Submitting a sitemap tells Google where the file lives. It does not guarantee every URL is indexed, and you should not refresh the report every hour.

4. Live URL test + request indexing

Open the URL Inspection tool (search bar at the top of Search Console).
Paste a full page URL that belongs to the verified property (for example your homepage or a new post).
Read the indexed result. If Google has never seen it, or the live page differs, click Test live URL.
Fix anything obvious: noindex, robots block, 404, soft redirect, wrong canonical.
When the live test looks healthy, click Request indexing. There is a daily limit — for many new pages, a sitemap with accurate lastmod beats spamming Request indexing.

Indexing often takes about a day, sometimes longer. A successful request is a queue ticket, not a promise.

Google Analytics 4 — setup (start-here, traffic)

GA4 standard is free. This site uses GA4 via a single gtag snippet in <head> — same pattern you should use on a static or Vercel site. Steps below follow Google’s setup help: create account / property / stream and install the Google tag. Skip Google Tag Manager unless you already live in GTM.

1. Create an Analytics account (first time only)

Go to analytics.google.com and sign in.
If you have never used Analytics, click Start measuring. Otherwise open Admin → Create → Account.
Name the account (your project or company). Set data-sharing options, then continue to the first property.

2. Create a GA4 property

In Admin → Create → Property (or continue from the account wizard).
Enter a property name, reporting time zone, and currency.
Pick industry and business size, then how you intend to use Analytics (this only shapes default reports).
Click Create and accept the Terms of Service / Data Processing Amendment when prompted.

3. Add a Web data stream

Admin → Data collection and modification → Data Streams → Add stream → Web.
Enter your site URL (for example https://example.com) and a stream name.
Leave Enhanced measurement on unless you know you want to turn events off later.
Click Create stream.
Copy the Measurement ID — it looks like G-XXXXXXXX. Treat it as public-by-design (it ships in page HTML), but still do not invent a second ID on the same page.

4. Install gtag once in `<head>`

On the stream page: Google tag → View tag instructions → Install manually.
Copy the full snippet Google shows (loads gtag/js?id=G-… then gtag('config', …)).
Paste it immediately after the opening <head> on every page (or once in your shared layout / HTML chrome). One Google tag per page — not a second copy in the body, and not “GA plus GTM plus another G-” stack on day one.
Deploy. Data collection can take up to ~30 minutes to start; then open the Realtime report and hit your own site to confirm hits.

Shape of the tag (your real ID replaces the placeholder):

<!-- Google tag (gtag.js) — paste after <head> -->
<script async src="https://www.googletagmanager.com/gtag/js?id=G-XXXXXXXX"></script>
<script>
  window.dataLayer = window.dataLayer || [];
  function gtag(){dataLayer.push(arguments);}
  gtag('js', new Date());
  gtag('config', 'G-XXXXXXXX');
</script>

If you sell to the EU, you still need a consent path before firing analytics cookies — this is not legal advice; it is why cookie banners exist next to GA.

5. Optional: link Search Console ↔ GA4

Once both products are live, you can link them in GA Admin so Search Console queries show up next to Analytics reports. Nice-to-have after the two setups work on their own.

Plausible / Umami

Self-host Plausible or Umami if you do not want Google. Plausible Cloud is a paid product. If you already have GA4, do not add these “for comparison” on production.

PostHog

PostHog Free (no card): 1 million analytics events, 5,000 session recordings, 100,000 error exceptions, 1 project, ~1-year retention. That is product analytics and a Sentry-shaped errors bucket. Pay-as-you-go after the include. Use it when you care about funnels, not only sessions.

Create a project → project API key in .env (POSTHOG_KEY / host).
Init the SDK once. Autocapture is noisy — name the events that matter.

Sentry

Sentry Developer: $0, 1 user, 5,000 errors, 5 GB logs, 5M spans, 50 replays, 1 uptime monitor, 30-day lookback. Team from $26/mo (annual default). This is the error specialist. SENTRY_DSN in .env — DSN is designed to live in the client SDK; still don’t commit extra tokens.

Better Stack

Better Stack Free: 10 monitors, 1 status page, 100,000 exceptions/month, 5,000 replays, 3 GB logs with 3-day retention. Official copy: personal projects. Use it when you want uptime + errors in one dashboard without Sentry’s 5k error ceiling.

Axiom

Axiom Personal: 500 GB/month data loading, 10 GB-hours query, 25 GB storage, 30-day retention, no card. Cloud is $25/mo with a larger include. Logs, not a GA replacement.

UptimeRobot

UptimeRobot Free: 50 monitors, 5-minute interval, hobby / non-profit positioning, no card. Solo is the paid step for 60-second checks. Ping / and the checkout API, not only the homepage.

Also: Microsoft Clarity

Clarity is free heatmaps and session recordings. One script next to GA if you want rage-clicks. It is not Search Console and not Sentry.

Free vs paid (the honest cliffs)

GSC / GA4 — stay free unless you buy GA360.
PostHog — card when you leave 1M events or want more projects.
Sentry — Team when a second human needs login, or 5k errors is a joke.
UptimeRobot / Better Stack — paid when the URL is a business (their Free copy is hobby/personal).

Practical tips

One analytics stack — GA4 or Plausible, plus optional PostHog for product. Not all three.
Errors ≠ uptime — keep a ping and an exception tracker.
Source maps on paid error tools — minified stacks waste the 5k Sentry include.
Quotas reset monthly — screenshot Usage after a launch spike.

What's next in Free vibe coding stack

Part 8 — payments — Stripe Checkout

Earlier: Part 6 — email & SMS.

FAQ

What is the difference between Google Search Console and Google Analytics?

Search Console shows how your site appears in Google Search (queries, clicks, indexing). Google Analytics (GA4) shows what visitors do after they land. Use both: Domain property + DNS verify + sitemap in Search Console, then a GA4 web stream + one gtag snippet in the HTML head for traffic. As of Sep 2026 both stay free on the standard tiers this guide uses.

What is the best free analytics stack for a vibe-coded side project?

Start with Search Console + GA4 for SEO and traffic, then add one product/error tool — usually PostHog Free (1M events, 100k errors/month) or Sentry Developer (5k errors, 1 user). Do not run GA4 and Plausible and PostHog on the same page.

How do I add free analytics to a vibe-coded project?

Create a GA4 property and web stream, paste one gtag snippet in the HTML head, verify Search Console (Domain + DNS), submit the sitemap, then use URL Inspection + Request indexing for key pages.

Is PostHog free? How does it compare to Google Analytics?

Yes for most side projects: PostHog’s free allowance includes about 1M analytics events per month (plus replay and error quotas in-product). GA4 is better for marketing attribution and Google Ads; PostHog is better for product events, funnels, replays, and bundled error tracking. Pick one traffic stack — GA4 or Plausible — and treat PostHog as the product layer if you need it.

PostHog vs Sentry for error tracking — which should I use?

PostHog Free includes a large error bucket (~100k exceptions/month) next to product analytics. Sentry Developer is free for 5,000 errors and one user, with deeper engineering tooling when you outgrow that. Many vibe projects start on PostHog alone; add Sentry when you want specialized error workflows.

Is Sentry enough without uptime monitoring?

No. Error trackers only see failures your running app reports. If the deploy never boots, DNS breaks, or the host is down, Sentry stays quiet. Pair errors with an external ping tool such as UptimeRobot Free (50 monitors, 5-minute checks) or Better Stack Free monitors.

Do I need a credit card for free monitoring and analytics tools?

No for the start-here stack in this guide: Search Console, GA4 standard, PostHog Free, Sentry Developer, Axiom Personal, Microsoft Clarity, and the free uptime tiers (UptimeRobot / Better Stack personal). You need billing only when you exceed free quotas or move to paid team plans (for example Sentry Team from about $26/mo).

Can I use Google Analytics and Plausible together?

You can technically load both scripts, but you should not for a small project. You double consent complexity, split the same traffic into two dashboards, and still miss product/error visibility. Choose GA4 or a privacy-first tool, then add PostHog only if you need product analytics.

Does Microsoft Clarity replace Google Analytics?

No. Clarity is free heatmaps and session recordings — a sibling to GA, not a full traffic analytics suite. Keep GA4 (or Plausible/Umami) for who visited and from where; add Clarity as one extra script if you want to watch rage clicks and scrolls.

Originally published on Vibe Coder's Life. Tuesday/Friday digest on AI coding tools — subscribe.

Sources: Search Console / GA4 help docs; PostHog, Sentry, Better Stack, Axiom, UptimeRobot, Plausible, Umami, Clarity pricing pages. Quotas checked 21 Sep 2026 — verify in-product.

Disclosure: All tools in this article are plain product URLs. No paid placement. This site uses Google Analytics 4 and Search Console. Drafting/editing used AI assistance.

Free email and SMS for vibe coding (2026)

Viktor Kondas — Mon, 12 Oct 2026 11:00:00 +0000

How to send transactional mail and OTP texts from a vibe-coded project — Resend as the start-here, then Brevo, SES, Loops, Mailtrap, and Twilio for SMS.

Free vibe coding stack — **Part 6* (email & SMS).*

Part 5 gave you login. Next up is a real sender — magic links, receipts, “your download is ready.” Part 6 is email, with a short SMS row for OTP when you need it.

I run this site’s contact form, newsletter digest, and Playbook fulfillment on Resend (news@ / hello@vibecoderslife.com). Same path I’d hand you in Cursor: one API key, verify the domain, send from it.

Skip Gmail SMTP for product mail. And Resend has no SMS — Twilio is that separate row when you need OTP texts, not a second email API.

Start-here: Transactional + newsletter → Resend (3,000/mo, 100/day, 3 domains on Free). Staging inbox → Mailtrap sandbox (tiny on Free). Marketing audience → Loops. OTP texts → Twilio after trial credit.

Email map

Provider	Best for	Free quota	First paid	Card?
Resend	Transactional API + digest	3,000/mo; 100/day; 3 domains	Pro $20/mo — 50k, no daily cap, 10 domains	No
Brevo	Higher daily free send	300 emails/day	Starter from $9/mo — 5k/mo, no daily cap	No
AWS SES	Cheap at scale	À la carte ~$0.10 / 1k (sandbox first)	No monthly floor — Essentials from ~$0.16 / 1k	AWS account
Loops	SaaS lifecycle / audience	1,000 contacts; 4,000 sends / 30 days	From $49/mo — up to 5k contacts, unlimited sends	No
Mailtrap	Staging / inspect mail	Sandbox: 50 tests/mo; API: 4k/mo, 150/day	API Basic from $15/mo (10k); Sandbox Basic from $17/mo (500 tests)	No

Checked 16 Sep 2026 from official docs. Resend Marketing is a separate 1,000-contact Free plan — not the transactional 3,000-send meter.

SMS (short row)

Provider	Best for	Free quota	First paid	Card?
Twilio	OTP / Verify / SMS	Trial credit	Pay-as-you-go — SMS from ~$0.0083; Verify from $0.05	No to start

Country rates vary. Do not send marketing SMS from a trial number and expect it to land. Checked 16 Sep 2026.

Wire it from Cursor

Wire transactional email to Resend using RESEND_API_KEY from .env.
Walk me through: (1) verifying the sending domain (SPF/DKIM),
(2) putting RESEND_API_KEY in .env and .gitignore,
(3) the same env var on the host dashboard,
(4) one contact or magic-link send.
Do not use Gmail SMTP. Do not add a second mail vendor in production.
If we need SMS OTP, add Twilio Verify in a separate change — Resend has no SMS.

Resend (start-here)

Resend Free: 3,000 emails/month, 100 per day, 3 domains, 10,000 automation runs, 30-day retention, ticket support. First paid is Pro at $20/mo: 50,000 emails, no daily cap, 10 domains, 5 webhook endpoints, 100 AI credits/mo. Scale starts higher if you need more volume or dedicated Slack. This is the DX start-here: Node SDK, React Email, webhooks, inbound.

Create an API key → RESEND_API_KEY in .env.
Add the domain → copy DNS (SPF, DKIM, DMARC). Wait for verify.
Send from that domain only. Rotate the key if it hits GitHub.

Brevo

Brevo Free: 300 emails/day (campaigns + transactional) after account approval. Unused daily quota dies at midnight. Branding on Free. First paid is Starter from $9/mo for 5,000 emails/month with no daily cap (higher send tiers cost more; removing the Brevo logo is often an add-on). Pick Brevo when Resend’s 100/day cap is the problem and you are fine with marketing-suite UX.

AWS SES

SES has no classic “first paid tier” — you pay for volume. À la carte outbound is about $0.10 per 1,000 emails; the newer Essentials plan lists about $0.16 per 1,000 for the first 10M/mo band. New accounts still start in the sandbox (verified addresses only) until AWS grants production access. New AWS accounts may get Free Tier credits that can apply to SES — check your Billing dashboard. Use SES when you want cheap scale and will accept IAM/sandbox toil.

Loops

Loops Free: 1,000 subscribed contacts, 4,000 sends in any rolling 30 days, 10 emails/sec, “Powered by Loops” footer. First paid starts around $49/mo for up to 5,000 subscribed contacts: unlimited sends, no branding, transactional included. That is an audience / lifecycle tool. Keep Resend for “forgot password” if Loops is only the product newsletter — or commit to Loops for both, not both vendors firing the same receipt.

Mailtrap

Mailtrap splits products. Email Sandbox Free: 50 test emails/month, 1 sandbox, 10 messages retained — enough to see the HTML, not to load-test. Sandbox Basic starts about $17/mo ($14/mo annual) for 500 tests. Email API Free: 4,000/mo, 150/day, 1 domain. API Basic from $15/mo for 10,000 emails, 5 domains, 3 users, no daily cap. Point preview/staging at sandbox SMTP; production stays Resend.

Twilio SMS

Twilio: sign up, trial credit, no card required to start. There is no fixed “first paid plan” — after trial you pay as you go. SMS starts around $0.0083 to send or receive (US list; HU/PL differ). Verify starts at $0.05 per verification. Use it for OTP, not as the mail provider. Env: TWILIO_ACCOUNT_SID, TWILIO_AUTH_TOKEN, Verify service SID.

Free vs paid (the honest cliffs)

Resend — Pro ($20/mo) when you blow 100/day (a small list on send day) or 3k/month — unlocks 50k and drops the daily cap.
Brevo — Starter ($9/mo) when 300/day or branding is the issue; you buy monthly send volume instead.
Loops — paid (~$49/mo) when you pass 1,000 subscribed contacts or need unlimited sends without the footer.
Mailtrap — Sandbox Basic when 50 tests/mo is too small; API Basic when Free API caps bite (keep production on Resend).
SES — production access + IAM toil; cheap volume, no friendly DX start-here.
Twilio — trial ends; SMS/Verify are never “unlimited free.”

Practical tips

DNS first — unverified domains go to spam. Wait for Resend’s green checks.
Mailtrap in preview, Resend in prod — same code path, different env.
Do not double-send — Loops or Resend for broadcasts, not both.
SMS is a separate product — consent, sender IDs, and country rates.

What's next in Free vibe coding stack

Part 7 — monitoring — GSC + GA4
Part 8 — payments — Stripe Checkout

Browse tools we actually use on vibecoderslife.com/tools. Resend is listed there as a plain URL.

FAQ

How do I add free transactional email?

Create a Resend API key, verify the domain, put RESEND_API_KEY in .env and on the host, and paste the Cursor prompt above.

Resend vs Brevo vs Loops?

Resend for API-first transactional (what I use on this site). Brevo for a higher daily free send inside a marketing suite. Loops for contact-based SaaS lifecycle. Pick one production sender.

Does Resend send SMS?

No. Use Twilio Verify / SMS. Keep Resend for email.

Can I use Gmail for the product?

Not if you care about deliverability or App Passwords breaking. Use Resend (or SES at scale).

Do I need a credit card?

Not for Resend Free, Brevo Free, Loops Free, Mailtrap Free, or Twilio trial. AWS and paid upgrades need billing.

Originally published on Vibe Coder's Life. Tuesday/Friday digest on AI coding tools — subscribe.

Sources: Resend pricing; Brevo pricing; AWS SES pricing; Loops pricing; Mailtrap pricing; Twilio pricing. Numbers as of 16 Sep 2026.

Disclosure: Resend, Brevo, AWS, Loops, Mailtrap, and Twilio are plain product URLs. No paid placement by those providers for this article. This site uses Resend in production. Drafting/editing used AI assistance.

Free auth for vibe coding (2026)

Viktor Kondas — Fri, 09 Oct 2026 11:00:00 +0000

How to add free user login to a vibe-coded project — Clerk as the start-here, then WorkOS, Supabase Auth, Kinde, and OSS (Better Auth / Auth.js).

Free vibe coding stack — **Part 5* (auth).*

Part 1 of Free vibe coding stack covered free cloud AI API keys. Part 2 was courses. Part 3 was hosting. Part 4 was database and storage. A Neon database without login is an open spreadsheet. Part 5 is auth — sessions you can ship without paying day one.

This chapter is user login. Email (magic links, receipts) is Part 6 — coming next in this series. If you picked Supabase in Part 4 because it bundles Auth, stay on Supabase Auth — do not add Clerk on top of the same app.

Start-here: Need hosted login Cursor actually wires → Clerk (Hobby: 50k monthly retained users per app). Quota is the constraint → WorkOS AuthKit (first 1M MAU free; production still wants a card on file). Already on Supabase → Supabase Auth (50k MAU on Free). OSS → Better Auth or Auth.js.

Hosted auth map

Provider	Best for	Free quota	Card?	Cliff
Clerk	Hosted UI + Next middleware	50,000 MRU / app (Hobby)	No	MRU ≠ MAU; MFA / passkeys are Pro ($25/mo)
WorkOS AuthKit	Bigger free MAU	First 1M MAU / month	For production	Payment method on file; custom domain $99/mo
Supabase Auth	Already on Supabase	50,000 MAU on Free	No	Pauses with the Free project after 1 week idle
Kinde	Auth + optional billing	10,500 MAU; 5 MAO	No	Smaller MAU than Clerk/WorkOS; Pro $25/mo

Copywriter-verified 3 Sep 2026; re-checked 8 Sep 2026 from official docs. Clerk bills monthly retained users (return ≥24h after signup). WorkOS and Supabase bill MAU. Quotas move — screenshot Usage the week you launch.

OSS map

Provider	Best for	Free quota	Card?	Cliff
Better Auth	Self-host, framework-agnostic	Your server	No	You own sessions, email, password resets
Auth.js / NextAuth	Next-native sessions	Your server	No	Same: you wire Resend (Part 6) yourself

Skip Auth0 and Firebase as primary in this chapter — leftover BaaS. If the agent scaffolded them anyway, one vendor, not two.

Wire it from Cursor

Same loop as Part 4: create the project, copy keys, put them in .env, add the same names on Vercel / Cloudflare / Netlify (Part 3). Never commit .env. Secret keys stay on the server. Paste this if you want the agent to stay on one vendor:

Wire this app to Clerk using NEXT_PUBLIC_CLERK_PUBLISHABLE_KEY and CLERK_SECRET_KEY from .env.
Walk me through: (1) creating the Clerk application if needed,
(2) putting keys in .env and .gitignore,
(3) the same env vars on the host dashboard,
(4) sign-in / sign-up / a protected route.
Do not add a second auth vendor. Do not put CLERK_SECRET_KEY in client code.
If this app already uses Supabase Auth, stop and keep Supabase — do not add Clerk.

Clerk (start-here)

Clerk is hosted sign-in UI, sessions, and Next.js middleware. Official Hobby is free: 50,000 monthly retained users (MRU) per application, unlimited apps, up to 3 dashboard seats, up to 3 social connections, a fixed 7-day session, and 1-day application-log retention. Custom domain is included on Hobby. A user counts as retained if they come back at least 24 hours after signup — one-time tire-kickers do not fill the meter (“First Day Free”). Pro is $25/mo ($20/mo annual). MFA, production passkeys, remove Clerk branding, and custom session lifetime sit on Pro, not Hobby. Hit the Hobby MRU limit and you get a one-month grace, then you upgrade — Hobby is a hard limit, not pay-as-you-go overage.

This series still starts with Clerk (DX, components, Cursor-friendly) the same way Part 4 starts with Neon, not the cheapest disk. WorkOS wins if the number on the pricing page is the only decision.

Create an application at dashboard.clerk.com → API keys.
Local: NEXT_PUBLIC_CLERK_PUBLISHABLE_KEY and CLERK_SECRET_KEY in .env. Host: same names.
Add the Clerk provider + middleware. Restrict CLERK_SECRET_KEY to server routes.

WorkOS AuthKit

WorkOS AuthKit is user management with a 1 million MAU free band, then $2,500/mo per extra million. Staging is free. Production AuthKit still requires a payment method on file even if you never cross 1M MAU. Hosted custom domain is $99/mo. Enterprise SSO / Directory Sync start around $125/connection — that is usually the first bill, not MAU.

Create an environment → copy WORKOS_API_KEY and WORKOS_CLIENT_ID.
Use AuthKit’s hosted UI or the SDK. Keep staging keys out of production.

Supabase Auth

If Part 4 already gave you a Supabase project, Auth is included: 50,000 MAU on Free, social providers, basic MFA. The same 7-day pause applies. Copy the anon key and URL you already have — do not invent Clerk “for nicer UI” beside it.

Authentication → providers in the Supabase dashboard.
SUPABASE_URL + SUPABASE_ANON_KEY in .env (service role on the server only).

Kinde

Kinde Free: 10,500 MAU, 5 monthly active organizations, unlimited apps, no card. MFA is included on Free. Pro from $25/mo. Users you bill through Kinde Billing (more than USD $4/month) do not count toward MAU. Smaller free band than Clerk or WorkOS — pick it if you also want Kinde’s billing, not as a second auth next to Clerk. Note vs Clerk: Kinde Free includes MFA; Clerk Hobby does not (MFA is Pro).

Create an app → KINDE_CLIENT_ID, KINDE_CLIENT_SECRET, KINDE_ISSUER_URL.
Set callback URLs for local and the Part 3 host.

Better Auth

Better Auth is TypeScript auth you run. No MAU meter. You store sessions (usually on the Part 4 database) and send magic-link / reset mail via Resend in Part 6 (email & SMS — next in the series). BETTER_AUTH_SECRET and BETTER_AUTH_URL in .env.

Auth.js / NextAuth

Auth.js is the Next-native OSS path. Same deal: you own providers, the database adapter, and email. AUTH_SECRET plus OAuth client IDs. Fine if the scaffold is already NextAuth — do not migrate to Clerk “because the blog said so” mid-demo.

Free vs paid (the honest cliffs)

Clerk — Hobby is a hard MRU cap. Pro when you need MFA / passkeys, or you outgrow 50k retained users.
WorkOS — AuthKit MAU is huge; the bill is production extras (custom domain, SSO connections).
Supabase — Auth is “free” until the project pauses or you need Pro for always-on.
Kinde — 10.5k MAU is the hobby ceiling unless billed users offset it.
OSS — $0 vendor fee; you pay in email deliverability and session bugs.

Practical tips

One auth vendor — Clerk or WorkOS or Supabase Auth or OSS. Two session cookies is how demos rot.
Secrets on the server — publishable / anon keys in the browser; secret / service role never.
Callback URLs — localhost plus the Part 3 production domain, both listed in the dashboard.
Quotas move — Clerk’s unit is MRU as of Feb 2026. Screenshot Usage.

What's next in Free vibe coding stack

You can log a user in. They still need mail, visibility, and (maybe) a card:

Part 6 — email & SMS — Resend start-here (coming next)
Part 7 — monitoring — GSC + GA4 (later in the series)
Part 8 — payments — Stripe Checkout (series closer)

Browse tools we actually use on vibecoderslife.com/tools (Cursor, Vercel, GitHub — affiliate links marked where active). Clerk is a plain product URL.

FAQ

How do I add free login to a vibe-coded project?

Create a Clerk application, put the publishable + secret keys in .env and on your host, and paste the Cursor prompt above. Do not add a second auth vendor “just in case.”

Clerk vs WorkOS?

Clerk if you want components and Next middleware (this series’ start-here). WorkOS if you need ~1M free MAU and will file a card for production. They are not the same billing unit (MRU vs MAU).

I already use Supabase. Should I add Clerk?

No. Use Supabase Auth. Clerk on top of Supabase is two user tables.

Is Better Auth really free?

The library is. Hosting, Postgres (Part 4), and transactional email (Part 6) are still your problem.

Do I need a credit card?

Not for Clerk Hobby, Supabase Free, Kinde Free, or OSS. WorkOS wants billing info before production. Clerk Pro / Kinde Pro / WorkOS add-ons need a card.

Originally published on Vibe Coder's Life. Tuesday/Friday digest on AI coding tools — subscribe.

Sources: Clerk pricing / Feb 2026 plans; WorkOS pricing / AuthKit environments; Supabase pricing; Kinde pricing; Better Auth; Auth.js. Copywriter-verified 3 Sep 2026; re-checked 8 Sep 2026.

Disclosure: Clerk, WorkOS, Supabase, Kinde, Better Auth, and Auth.js are plain product URLs. No paid placement by those providers for this article. Drafting/editing used AI assistance; the auth map is from my own shipping work.

AllStackd is live: one workspace for every Amazon AWS SES account

Viktor Kondas — Thu, 08 Oct 2026 11:00:00 +0000

I run a lot of domains on Amazon SES, so I built the control plane I wanted. Monitor every SES account, region and domain in one place, keep delivery in your own AWS account, and send transactional mail through an optional durable API. 14-day trial, no card.

Back in July I wrote about the AWS AI Get Certified Challenge for Central and Eastern Europe. I got in. The first month of Skill Builder is done — online training and the digital classroom. Live sessions start in October and run through the end of the year as I keep going on the AWS AI Practitioner path.

At some point reading about IAM and shared responsibility isn’t enough. I wanted to put what I’d learned so far into practice — on a real product, with real domains, in my own AWS account.

What is Amazon SES? SES stands for Simple Email Service. It’s Amazon’s API for sending transactional email — welcome messages, password resets, receipts — from your own AWS account. Pay-as-you-go, no contract. Great at delivery. The hard part is operating it when you have more than one product, region or domain.

Here’s the result.

AllStackd went live on 30 Sep 2026. It’s an Amazon SES control plane for every product. You monitor every SES account, region and domain from one workspace. Delivery stays in your own AWS account. And when you want it, there’s an optional durable API for transactional mail.

AllStackd is not an email host and it doesn’t replace SES. AWS delivers the mail and bills the sends. AllStackd bills the workspace.

Start here: allstackd.com → Start 14-day trial (no card) → create an AWS connection in Settings → upload the CloudFormation template in your AWS account → paste back the three stack outputs. Setup guide: allstackd.com/docs.

Why I built it

I have a lot of domains. Side projects, experiments, sites that need a welcome email, a password reset or a receipt. I was already deep in the AWS CEE learning track, so Amazon SES was the obvious place for their mail: pay-as-you-go, no contract, and delivery in an account I own.

SES is great at sending. The pain is everything around it:

Sandbox status is per AWS Region. A new SES account starts in the sandbox, where you can only mail verified addresses, max 200 messages per 24 hours and 1 per second. Production in eu-central-1 doesn’t mean production in us-east-1.
DKIM breaks quietly. Someone edits DNS, one of the three Easy DKIM CNAMEs disappears, and you find out when a user says they never got the email.
The console is one account at a time. The SES account dashboard and reputation page show health for one AWS account. With several products and regions, “check every account by hand” becomes the risk.

The AWS AI Practitioner program isn’t an email course. But it expects familiarity with IAM, the shared responsibility model and AWS pricing models. Those three ideas are exactly what AllStackd is built on: a scoped IAM role instead of keys, your account stays yours, and AWS bills AWS usage.

So I built the layer I wanted for my own domains, then shaped it for other indie developers like me.

What AllStackd is

AllStackd operates the layer above SES. Your AWS account remains the delivery system. AllStackd is the health view, the policy gate, and an optional sending API.

It has two operating modes:

Mode	What it does
Observe	See SES identities, DKIM, SPF, DMARC, sandbox status, quotas and reputation without routing email through AllStackd. Read-only against SES configuration and reputation APIs.
Control	Add a durable, policy-enforced sending layer project by project, with idempotency, suppressions and short retention. Delivery still happens in your AWS account.

Health findings come with a next safe action: DKIM drift, sandbox and quota issues surface in the dashboard before the next release.

Observe mode: health across SES connections — Healthy, DKIM drift, Sandbox — plus a next safe action.

What the site compares it to:

Amazon SES alone	SES + AllStackd
Operate accounts and regions separately	One workspace across SES connections
Manual identity and DKIM checks	Health findings with a next safe action
App-specific SES credentials	Project-scoped API keys
DIY queue and retry logic	Durable Control-mode sending workflow

How it works

Setup takes four steps.

AWS + SES. Verify a domain you own in SES (Easy DKIM CNAMEs, plus SPF and DMARC). Request production access when you’re ready to mail real users. AWS reviews that request, not AllStackd, and AWS says it gives an initial response within 24 hours.
CloudFormation role. In AllStackd → Settings, create a connection and download the template. In AWS CloudFormation: Create stack → With new resources → Upload a template file. Paste the RoleArn, SnsTopicArn and ConfigurationSetName outputs back into AllStackd, then confirm the SNS subscription.
Domain health. Confirm DKIM, SPF and DMARC on the Domains page. Owned domains only, no Gmail or Outlook as the From domain.
Optional Control API. Create a project, assign the connection, mint an as_ API key and send with an idempotency key.

Docs: prepare SES in your AWS account, then connect with CloudFormation — not long-lived keys.

A send looks like this:

curl -X POST https://allstackd.com/api/v1/emails \
  -H "Authorization: Bearer $ALLSTACKD_API_KEY" \
  -H "Idempotency-Key: welcome-user-42" \
  -H "Content-Type: application/json" \
  -d '{
    "from":"hello@your-verified-domain.com",
    "to":["customer@example.com"],
    "template":{"alias":"welcome","variables":{"first_name":"Ada"}}
  }'

A few details I care about as a builder:

202 means queued, not delivered. Poll GET /api/v1/emails/:id or use a signed webhook (HMAC-SHA256) for send, delivery, bounce, complaint and reject events.
Templates start from starters (welcome, verify-email, password-reset, magic-link, receipt) or raw HTML, with {{variables}} so copy can change without a deploy.
Up to 50 recipients per request, and retries with the same Idempotency-Key are safe.
TypeScript SDK file you copy into your app, plus an OpenAPI spec at /openapi.json.
MCP beta at https://allstackd.com/api/mcp: your coding agent can list domains, manage templates and queue a send with the same scoped API key.

What stays in your AWS account

This was the part I wouldn’t compromise on.

No long-lived access keys, no SES SMTP passwords. The CloudFormation stack creates an IAM role that trusts only AllStackd’s runtime role, conditioned on a unique External ID per connection. Sessions last at most one hour.
Observe is read-only. The permission list is published on allstackd.com/security, and you can review the template before you create the stack.
Control is honest about scope. Today it includes ses:SendEmail plus identity and configuration-set management. Narrowing send by configuration set or From address is on the hardening roadmap. If you only need health, start with Observe.
Message content is short-lived. In Control mode, payloads are encrypted at rest (AES-256-GCM) while queued and deleted within 24 hours. API keys are hashed and shown once.
If AllStackd is down, mail SES already accepted keeps delivering in your account. Control-mode sends wait in the durable queue and are retried.

Pricing

Every plan includes Observe and Control. An SES connection is one AWS account in one region. Recipients/month count Control-mode recipients submitted through AllStackd; Observe doesn’t meter mail you send outside AllStackd.

Plan	Monthly	Yearly	SES connections	Control recipients/month	Signed webhooks
Portfolio	€9	€89	1	100k	1
Studio (best fit)	€29	€289	5	500k	5
Agency	€79	€789	25	2m	20

14-day trial, no card during the trial. After 14 days, checkout is required to keep access.
Agency is one operator managing many client connections. Team seats aren’t part of this beta.
Prices exclude Amazon SES / AWS charges and may exclude tax. Signup is open worldwide.
AllStackd is a paid beta: there’s no SLA yet.

AWS bills the sending separately. As of Sep 2026, new SES accounts start on the Essentials plan at $0.16 per 1,000 emails (first 10M/month), and you can switch to à-la-carte at $0.10 per 1,000 outbound (Amazon SES pricing).

Who it’s for

Founders and studios: track identities, quotas, sandbox state and reputation across the products you operate.
Agencies: separate client AWS access and see health issues before campaigns or releases.
Freelance developers: keep each client’s SES account apart and catch DKIM or sandbox issues first.
SaaS teams: watch quotas, bounces and domain health for the transactional mail your product already sends.

And who doesn’t need it: one SES account, one product, low volume — stay in the Amazon SES console. I mean that. AllStackd fits when you run several products, regions or client AWS accounts.

For me, that’s the vibe coder with a handful of domains and several side projects, each needing its own mail. I’m a solo developer in Hungary, and AllStackd is built for people who work the same way.

Running more than one product on SES? Start the 14-day trial (no card) and connect one account in Observe mode first. Founder-assisted setup is available → allstackd.com.

Your turn — get started

Verify a domain in Amazon SES — Easy DKIM, SPF and DMARC on a domain you own.
Request production access — for that region when you’re ready for real recipients. Sandbox is per region.
Start the 14-day trial — sign up at allstackd.com. No card during the trial.
Connect with CloudFormation — Settings → create a connection → download the template → create the stack in your SES region → paste RoleArn, SnsTopicArn, ConfigurationSetName → confirm the SNS subscription.
Check Domains health — confirm DKIM / SPF / DMARC in Observe mode.
Optional Control send — create a project, mint an as_ key, send a test email.
Stuck or found a bug? — read allstackd.com/docs, email support@allstackd.com for founder-assisted setup, or reply to the newsletter.

If you only remember one line: SES delivers. AllStackd operates.

FAQ

What is an Amazon SES dashboard?

The built-in one is the SES console's account dashboard: sending activity, quota used, bounces and complaints for a single AWS account, with sandbox status per region. A third-party SES dashboard like AllStackd sits on top of that and pulls identities, DKIM, SPF, DMARC, sandbox status, quotas and reputation from several SES accounts and regions into one workspace, while delivery stays in your AWS account.

How do I monitor multiple Amazon SES accounts in one place?

Connect each AWS account and region to one monitoring layer instead of logging into every console. In AllStackd, each connection is one AWS account in one region, added through a CloudFormation-created IAM role with a unique External ID, not access keys. Observe mode then shows identity health, DKIM drift, sandbox status, quotas and reputation for all connections in one view. Portfolio covers 1 connection, Studio 5, Agency 25.

Is AllStackd free?

No, but there's a 14-day free trial and no card is required during it. After the trial, checkout is required to keep access. Plans are Portfolio at €9/month or €89/year, Studio at €29/month or €289/year, and Agency at €79/month or €789/year. Every plan includes Observe and Control. Amazon SES sending charges are billed separately by AWS, and prices may exclude tax.

Does AllStackd replace Amazon SES?

No. AllStackd is not an email host. Messages are delivered by Amazon SES in your own AWS account, and AWS bills you for those sends. AllStackd is the operations layer on top: a health view across accounts and regions, plus an optional Control-mode API that queues transactional email with idempotency and suppressions before submitting it through your SES account.

Do I need to give AllStackd my AWS access keys?

No. You download a CloudFormation template from AllStackd, create the stack in your AWS account, and paste back the role, SNS topic and configuration set outputs. AllStackd assumes that role with a unique External ID, and sessions last at most one hour. It never asks for long-lived access keys or SES SMTP passwords. Observe mode is read-only; Control mode adds sending permissions.

How do I get out of the Amazon SES sandbox?

Request production access in the SES console: Account dashboard → View Get set up page → Request production access. Choose Transactional or Marketing, add your website URL, and confirm you only mail people who asked for it and handle bounces and complaints. AWS says it sends an initial response within 24 hours. Sandbox status is per region, so repeat this for every region you send from.

How much does Amazon SES cost in 2026?

As of Sep 2026, new SES accounts start on the Essentials plan at $0.16 per 1,000 emails for the first 10 million per month, with no monthly fee. À-la-carte outbound is $0.10 per 1,000, and attachments add $0.12 per GB. Pro ($105) and Enterprise ($500) plans add a per-account, per-region monthly fee. New AWS customers can get up to $200 in Free Tier credits.

Originally published on Vibe Coder's Life. Tuesday/Friday digest on AI coding tools — subscribe.

Sources: AllStackd · Amazon SES pricing · AWS Certified AI Practitioner.

Disclosure: AllStackd is my product (paid beta with a 14-day trial). No affiliate links in this post. Drafting/editing used AI assistance.

Free databases & storage for vibe coding (2026)

Viktor Kondas — Wed, 07 Oct 2026 11:00:00 +0000

How to add free Postgres, Redis, and file storage to a vibe-coded project — Neon as the start-here, then Supabase, Turso, D1, R2, and when the free tier ends.

Free vibe coding stack — **Part 4* (database & storage).*

Part 1 of Free vibe coding stack covered free cloud AI API keys. Part 2 was courses. Part 3 was hosting. A live URL without state is a brochure. Part 4 is database and storage — Postgres, Redis, and files you can attach without paying day one.

This chapter is persistence: the connection string and the free-tier cliffs. If you pick Supabase because it also bundles login, still stop at the database today — do not invent auth in this change.

Start-here: Need a real SQL database → Neon (serverless Postgres, scale-to-zero, DATABASE_URL). Need Auth + Storage + Realtime in one product → Supabase (still Postgres; Free projects pause after a week idle). Files next to Neon → Cloudflare R2 (no egress). Already on Vercel from Part 3 → Vercel Blob.

Database map

Provider	Best for	Free quota	Card?	Cliff
Neon	Serverless Postgres	100 projects; 0.5 GB + 100 CU-h / project / mo	No	Scale-to-zero after 5 min (cannot disable on Free)
Supabase	Postgres + extras	2 active projects; 500 MB DB; 1 GB files	No	Pauses after 1 week inactivity
Turso	SQLite / edge replicas	5 GB; 500M row reads; 10M writes / mo	No	Rows-read meter; Developer ~$5/mo
Cloudflare D1	SQLite next to Pages	5 GB account; 500 MB / DB; 5M reads/day	No	Hits Workers Free daily row caps
MongoDB Atlas M0	If the agent scaffolded Mongo	512 MB; 1 Free cluster / project	No	100 ops/sec; 10 GB in/out per 7 days
Upstash Redis	Cache / rate-limit / queue	256 MB; 500K commands / mo	No	Not a primary app database

Checked 18 Aug 2026 from official docs. Quotas move — screenshot Usage the week you launch.

Storage / CDN map

Provider	Best for	Free quota	Card?	Cliff
Cloudflare R2	Object storage next to Neon	10 GB-month; 1M Class A; 10M Class B	No	Egress $0 from R2
Vercel Blob	Files on a Vercel app	Included in Hobby envelope	No	Hobby cannot buy extra; Blob pauses at cap
Backblaze B2	S3-compatible backup	First 10 GB storage always free	No*	Egress free up to 3× stored (or via CDN partners)
UploadThing	Drop-in Next uploads	2 GB storage	No	100 GB plan is $10/mo
ImageKit	Image CDN / transforms	20 GB bandwidth / mo; 3 GB DAM	No	Delivery stops at the bandwidth cap
bunny.net	Cheap paid CDN	14-day trial	For paid	$1/mo minimum; ~$0.01/GB EU/NA — not forever-free

Cloudflare KV (not in the table): 1 GB stored, 100k reads/day on Workers Free — key/value next to D1/R2, not a SQL database. *B2 billing starts after the 10 GB include.

Wire it from Cursor

Same loop on every provider: create the project, copy the connection string or token, put it in .env, add the same names on Vercel / Cloudflare / Netlify (Part 3). Never commit .env. Paste this if you want the agent to stay on one database:

Wire this app to Neon Postgres using DATABASE_URL from .env.
Walk me through: (1) creating the Neon project if needed,
(2) putting DATABASE_URL in .env and .gitignore,
(3) the same env var on the host dashboard,
(4) a minimal schema/migration.
Do not add a second database. Do not invent Clerk/Auth in this change.
If we need file uploads, use Cloudflare R2 or Vercel Blob — not a second Postgres.

Neon Postgres (start-here)

Neon is serverless Postgres: compute scales to zero after 5 minutes idle on Free (you cannot turn that off). Idle prototypes accrue zero CU-hours. This site uses Neon. Official plans: Free is permanent, no card — 100 projects, 10 branches / project, 100 CU-hours / project / month, autoscale up to 2 CU, 0.5 GB storage / project, 5 GB public network transfer / month.

Honest: storage and CU-hours still cap. Hit CU-hours or egress and compute suspends until the next billing period (or you upgrade). Hit 0.5 GB and writes that grow storage fail. Launch is pay-as-you-go (no monthly minimum) if you outgrow Free.

Sign up at neon.com → New project → copy the pooled DATABASE_URL.
Local: DATABASE_URL=postgres://… in .env. Host: same name in the dashboard.
Optional: Neon MCP in Cursor if you want the agent to inspect schema without pasting dumps into chat.

Supabase

Supabase is Postgres plus Auth, Storage, Realtime, and Edge Functions. Free: 2 active projects, 500 MB database, 1 GB file storage, 5 GB egress, 50k MAU. Official: Free projects pause after 1 week of inactivity. Restore from the dashboard; data is not deleted. Pro starts at $25/mo and does not pause.

Pick Supabase when you actually want the bundle. Do not run Neon and Supabase for the same app. Today, copy DATABASE_URL or the anon/service keys into .env and stop — do not invent login in this change.

New project at supabase.com/dashboard.
Project Settings → Database → URI, or API keys for the JS client.
Expect a pause email if the side project goes quiet for a week.

Turso

Turso is libSQL (SQLite fork) with edge replicas. Free: 5 GB, 100 databases, 500 million rows read and 10 million rows written per month. Paid starts at Developer (~$5/mo). The bill that surprises people is rows read — an unindexed SELECT * can chew the meter. Use it for local-first / SQLite-shaped apps, not as a second Postgres next to Neon.

Create a DB in the Turso console → copy TURSO_DATABASE_URL and TURSO_AUTH_TOKEN.
Wire via @libsql/client or Drizzle. Index the columns you filter on.

Cloudflare D1

D1 is SQLite on Workers. If Part 3 put you on Cloudflare Pages, D1 is the in-family database. Workers Free: 5 GB storage total, 500 MB per database, 10 databases, 5 million rows read / day, 100,000 rows written / day. No D1 egress charge. Daily caps reset midnight UTC; queries fail when you blow them until the next day (or Workers Paid).

Dash → Workers & Pages → D1 → Create database.
Bind it in wrangler.toml (or Pages project settings). There is no classic DATABASE_URL — it is a Worker binding.

MongoDB Atlas

Atlas Free (M0): 512 MB, shared CPU/RAM, one Free cluster per project, ~100 ops/sec, 10 GB in and 10 GB out per rolling 7 days. Use it if Cursor already generated Mongoose. Do not add it “for flexibility” beside Neon.

Create a Free cluster → Database Access + Network Access (your IP or 0.0.0.0/0 for a demo, tighten later).
MONGODB_URI in .env.

Upstash Redis

Upstash Free: 256 MB, 500,000 commands / month, 10 GB bandwidth, one database. That is cache, rate limits, and queues — not your users table. Older “10k commands/day” blurbs are stale; use the monthly command cap.

Create a Redis DB → copy UPSTASH_REDIS_REST_URL and UPSTASH_REDIS_REST_TOKEN (or the Redis URL).
Keep TTL on cache keys so you do not fill 256 MB with junk.

Cloudflare R2

R2 is S3-compatible object storage with no egress from R2 itself. Free: 10 GB-month, 1 million Class A (writes/lists), 10 million Class B (reads). Default files pick if Neon is the database. Pair it with a public bucket + custom domain, or a Worker that signs URLs.

Dash → R2 → Create bucket. Create an API token with Object Read & Write.
Env: R2_ACCOUNT_ID, R2_ACCESS_KEY_ID, R2_SECRET_ACCESS_KEY, R2_BUCKET. Point the S3 client at https://<accountid>.r2.cloudflarestorage.com.

Vercel Blob

Vercel Blob is the low-friction file store if Part 3 left you on Vercel. Hobby includes Blob inside the shared usage envelope; Hobby cannot buy extra — when you hit the cap, Blob pauses until the window resets. Pro pays overages (and still has the commercial-use rule from Part 3). This site uses private Blob for a paid download zip.

Vercel dashboard → Storage → Blob → create store → copy BLOB_READ_WRITE_TOKEN.
Use @vercel/blob put() / head(). Dashboard browsing counts as advanced operations.

Backblaze B2

B2: first 10 GB of storage is always free. Egress is free up to 3× average monthly storage, or unlimited through CDN partners (Cloudflare, bunny.net, Fastly, and others). Good S3-compatible backup/archive next to R2, not the default public-image CDN.

Create a bucket → Application Key with access to that bucket.
Use the S3-compatible endpoint in the same way as R2, with B2’s region URL.

UploadThing

UploadThing is “better S3 for Next”: auth on your server, upload on theirs. Free: 2 GB storage shared across apps, unlimited uploads/downloads. 100 GB App is $10/mo.

Create an app → UPLOADTHING_TOKEN in .env.
Define a file router; keep the 2 GB include in mind for user-generated dumps.

ImageKit

ImageKit Forever Free: 20 GB bandwidth / month and 3 GB DAM storage. Delivery stops when you hit the bandwidth cap (it resets next calendar month). Lite is $9/mo with overages. Use it for transforms (resize, format) — not as a second object store if R2 already holds originals.

bunny.net

bunny.net is the cheap paid CDN: 14-day trial, then pay-as-you-go with a $1 / month minimum and about $0.01/GB in Europe & North America. It is not a free database and not a forever-free CDN. Reach for it when Cloudflare’s free static/R2 path is not the CDN you want, and you accept a small bill.

Also: Cloudflare KV

Workers KV on Free: 1 GB stored, 100,000 reads / day, 1,000 writes / day. Session flags, config, A/B — not relational data. Bind it like D1. No KV egress charge.

Free vs paid (the honest cliffs)

Free tiers are for hobby, demos, and learning. The first paid step is usually always-on compute or leaving the storage cap, not a fancy feature.

Neon — Launch when you need to disable scale-to-zero, more than 0.5 GB, or more CU-hours. Pay-as-you-go; no fake $25 floor.
Supabase — Pro at $25/mo removes the 7-day pause and raises disk/storage. That is the “this URL must stay up” tax.
D1 / R2 / KV — Workers Paid when daily row or request caps bite. R2 paid storage is cheap; egress stays $0 from R2.
Vercel Blob — Pro if Hobby usage pauses you, and if Part 3 already forced Pro for Stripe/commercial.
Redis — stay Free for cache; do not “upgrade Redis” to avoid designing a real database.

Practical tips

One primary database — Neon or Supabase or D1. Redis on the side is fine. Two SQL engines is how demos rot.
Env vars in the host dashboard — same names as local. Rotate if a URL hits GitHub.
Pooled Neon URL — serverless hosts want the pooled connection string, not a single long-lived client per isolate without a pooler.
Quotas move — screenshot Usage the week you launch.
Supabase pause — a quiet week is enough. Neon idle is cheap; Supabase idle can go to sleep.

What's next in Free vibe coding stack

Auth — free user login stacks (Part 5)
Email & SMS — transactional mail without a card day-one
Monitoring — Search Console, analytics, alerts
Payments — Checkout when Hobby is no longer enough

Browse tools we actually use on vibecoderslife.com/tools (Neon, Vercel, GitHub, Cursor — affiliate links marked where active).

FAQ

How do I add a free database to a vibe-coded project?

Create a Neon project, put the pooled DATABASE_URL in .env and on your host, and paste the Cursor prompt above. Do not add a second SQL database “just in case.”

Neon vs Supabase?

Neon if you only need Postgres (branching, scale-to-zero, this site’s stack). Supabase if you want Auth + file Storage + Realtime in one dashboard and you accept the 7-day pause on Free. Both are Postgres underneath.

Is Redis a free database for my app?

No. Upstash Redis is cache / rate-limit / queue. Store users and orders in Postgres (or SQLite/D1), not in 256 MB of Redis.

What about R2 vs Vercel Blob?

R2 if you want S3-compatible storage and zero egress. Blob if the app already lives on Vercel and you want a token + SDK with no extra vendor. Hobby Blob cannot buy extra usage.

Do I need a credit card?

Usually not for Neon, Supabase Free, Turso Free, D1/R2/KV on Workers Free, Atlas M0, Upstash Free, UploadThing Free, or ImageKit Forever Free. bunny.net wants a card after the trial. Paid upgrades (Neon Launch, Supabase Pro, Workers Paid) need billing.

Originally published on Vibe Coder's Life. Tuesday/Friday digest on AI coding tools — subscribe.

Sources: Neon plans / pricing; Supabase pricing / free project pausing; Turso pricing; D1 pricing; MongoDB Atlas; Upstash Redis; R2 pricing; Vercel Blob; Backblaze B2; UploadThing; ImageKit; bunny.net; KV pricing. Numbers as of 18 Aug 2026 — verify in-product.

Disclosure: Neon signup is a referral link (same as on our Tools page). Supabase, Turso, Cloudflare, MongoDB, Upstash, Vercel, Backblaze, UploadThing, ImageKit, and bunny.net are plain product URLs. No paid placement by those providers for this article. Drafting/editing used AI assistance; the stack map is from my own shipping work.

Free hosting for vibe coding (2026)

Viktor Kondas — Mon, 05 Oct 2026 11:13:59 +0000

How to host your vibe-coded project for free — Cloudflare Pages, Vercel Hobby, Netlify, GitHub Pages, and Render, plus when Hobby is not enough (Stripe / commercial).

Free vibe coding stack — **Part 3* (hosting).*

Part 1 of Free vibe coding stack covered free cloud AI API keys. Part 2 was courses. Keys and training still leave you on localhost:3000. Part 3 is hosting — a live HTTPS URL for the frontend you just vibe-coded.

This chapter is Git frontend hosts. Databases and file storage are next in this series — Neon is the start-here Postgres, not a website host. Auth comes after that.

Start-here: Static HTML / CSS / JS or a marketing page → Cloudflare Pages (static requests are free and unlimited) or GitHub Pages (simplest). Next / React from Cursor → Vercel Hobby for DX — but Hobby is non-commercial personal use. Stripe, ads, donations, or paid client work means Pro.

Provider map

Provider	Best for	Free quota	Card?	Commercial	Custom domain
Cloudflare Pages	Static / edge	Unlimited static requests; 500 builds/mo	No	OK for static hobby + business sites*	Yes (100 / project on Free)
Vercel Hobby	Next / React DX	100 GB Fast Data Transfer / mo	No	Personal only	Yes (50 / project)
Netlify Free	Jamstack / forms	300 credits / mo (new accounts)	No	OK until credits run out	Yes
GitHub Pages	Docs / portfolios	1 GB site; soft 100 GB bandwidth	No	Not for store / SaaS	Yes
Render static	Git → CDN frontend	Hobby: 5 GB bandwidth, 500 build min	No*	Hobby is personal / prototype	Yes (2 included on Hobby)

Checked 17 Aug 2026 from official docs. Quotas move — re-check the dashboard before you promise a client a $0 bill. *Cloudflare Functions share the Workers Free 100k req/day cap. *Render bills extra bandwidth if you add a card and exceed the include; without a card, Free services can suspend.

Deploy from GitHub in Cursor

Same loop on every host below: push the repo, connect the GitHub app, pick the branch, wait for a *.pages.dev / *.vercel.app / *.netlify.app / *.github.io / *.onrender.com URL. Paste this into Cursor if you want the agent to walk you through one host:

Help me deploy this project to Cloudflare Pages or Vercel.
Walk me through: (1) putting the repo on GitHub if it is not there,
(2) connecting the host, (3) the correct build command and output folder,
(4) env vars for any API keys (never commit .env),
(5) the live URL. Do not add Stripe or ads on a free Hobby plan.

Cloudflare Pages

Cloudflare Pages is the generous static pick. Official docs: requests to static assets are free and unlimited on free and paid plans. You get 500 builds per month, one concurrent build, 100 projects, 100 custom domains per project, 20,000 files, 25 MiB max file.

The "unlimited" headline stops at Functions. Pages Functions share your Workers Free quota — 100,000 requests per day, resetting midnight UTC. A brochure site that never invokes Functions stays in the unlimited bucket. An API-on-the-edge app does not.

Create a free Cloudflare account at dash.cloudflare.com.
Workers & Pages → Create → Pages → Connect Git (GitHub or GitLab).
Set the build command and output directory your framework expects (or leave empty for a plain HTML folder).
Deploy. Add a custom domain in project settings when you have one.

Vercel Hobby

Vercel is still the lowest-friction path for Next.js and most React apps Cursor scaffolds. Hobby is $0. Fair-use (as of 29 Jul 2026) includes about 100 GB Fast Data Transfer per month, 1M edge requests, 1M function invocations, 200 projects, 50 domains per project. Exceed a cap and that feature typically pauses until the window resets — Hobby cannot buy extra usage.

The binding constraint is not bandwidth. It is the commercial-use rule: Hobby teams are non-commercial personal use only. Pro is $20/mo. See Hobby vs paid below.

Sign up at vercel.com with GitHub.
Add New → Project → import the repo. Accept the detected framework.
Add env vars in Project → Settings → Environment Variables (same names as local .env).
Every push to the production branch ships. Pull requests get preview URLs.

Netlify

Netlify still fits Jamstack sites that want forms and deploy previews. New accounts since 4 Sep 2025 are on credit-based Free: 300 credits per month, hard limit, no auto-recharge. When credits hit zero, projects pause and visitors see "Site not available" until the next cycle or you upgrade. Bandwidth is 20 credits per GB; each production deploy is 15 credits; deploy previews are 0 credits.

Older (legacy) Free accounts may still show a 100 GB-style bandwidth include. Do not assume that if you signed up this year — check Usage & billing.

Sign up at app.netlify.com with GitHub.
Add new site → Import from Git → pick the repo and build settings.
Watch the credit ledger if you deploy to production often (15 credits a pop adds up).

GitHub Pages

GitHub Pages is static only — no serverless, no SSR. Official limits: published site ≤ 1 GB, recommended source repo 1 GB, deploy timeout 10 minutes, soft 100 GB bandwidth / month, soft 10 builds/hour unless you use Actions. Custom domains work. It is not allowed as free hosting for e-commerce, commercial SaaS, or sites whose main job is taking payments.

Push a public repo (or use Actions from a private repo on a paid GitHub plan if you already have one).
Settings → Pages → deploy from main / /docs or a GitHub Actions workflow.
Wait for https://<user>.github.io/<repo>/.

Render (static)

Render static sites are free to deploy: Git-connected, CDN, TLS, PR previews. They count against the Hobby workspace include: 5 GB outbound bandwidth / month (then $0.15/GB if you have billing set up), 500 pipeline minutes, 2 custom domains included.

Do not mix this up with Free web services. Those are a different product: they spin down after 15 minutes idle, take about a minute to wake, and eat 750 Free instance hours / month. Fine for a toy API. Bad if you tell someone your "free static site" will cold-start.

Sign up at dashboard.render.com.
New → Static Site (not Web Service) → connect the repo → build command + publish directory.
Confirm the URL ends up on the CDN path, not a sleeping Node process.

Hobby vs paid — Stripe is the trigger

Free tiers are for hobby, demos, and learning. If you need a payment provider on the live URL, plan on a paid hosting tier.

Vercel Hobby vs Pro is the worked example. Fair use says Hobby is personal / non-commercial. Commercial includes:

Any method of requesting or processing payment from visitors (Stripe Checkout counts)
Advertising the sale of a product or service
Getting paid to create, update, or host the site
Affiliate linking as the primary purpose of the site
Ads (including AdSense)
Asking for donations

There is no "under $10 MRR is still Hobby" clause. One paying customer, or even a donate button, is the policy line. Pro is $20 / month per deploying seat, with a usage credit and the ability to pay for overages. Hobby cannot buy extra bandwidth.

Same idea, other hosts:

Cloudflare — static can stay $0; Functions / Workers traffic can blow the 100k/day Free cap. Paid Workers starts around $5/mo.
Netlify — Free cannot buy credits. Personal is $9/mo (1,000 credits + recharge). Pro starts at $20/mo.
GitHub Pages — policy, not a price list: no store / SaaS on Pages.
Render — static is cheap until bandwidth; always-on backends are Starter web services (~$7/mo), not Free.

Also: Replit and Hostinger

These are not in the free Git CDN map. They solve different jobs.

Replit is build-and-host in the browser. If you never left Replit, publishing there is simpler than exporting to GitHub → Vercel. It is still a platform plan with usage limits — not unlimited Cloudflare-style static. Use it when the project already lives there.

Hostinger is cheap traditional hosting: cPanel, WordPress, shared PHP. Useful if someone wants a classic site builder, not a Git push. It is paid, not a free frontend CDN. Do not pick it as the default for a Cursor Next.js app.

Practical tips

Env vars in the dashboard — never commit .env. Rotate keys if they hit GitHub.
Custom domain + HTTPS — all five Git hosts issue certificates; DNS is the slow part (A / CNAME at your registrar).
Preview deploys — share the PR URL, not localhost screenshots.
Quotas move — screenshot Usage the week you launch; Netlify credits and Vercel fair-use are the two that surprise people.
One host per project — do not dual-deploy "for backup" until you know which URL is canonical.

What's next in Free vibe coding stack

Hosting is the public URL. Coming parts in this series:

Database & storage — free Postgres / files (Neon start-here)
Auth — free user login stacks
Email & SMS — transactional mail without a credit card day-one
Monitoring — Search Console, analytics, and alerts
Payments — Checkout when Hobby is no longer enough

Browse tools we actually use on vibecoderslife.com/tools (Vercel, Hostinger, Replit, GitHub, Cursor — affiliate links marked where active).

FAQ

How do I host a vibe-coded project for free?

Push the repo to GitHub, connect Cloudflare Pages (static) or Vercel Hobby (Next/React), and use the *.pages.dev or *.vercel.app URL. Stay off Stripe, ads, and paid client work while you are on Hobby-class plans.

Is Vercel Hobby enough if I add Stripe?

No. Vercel's fair-use rules treat payment processing as commercial. Upgrade to Pro (or host somewhere whose terms allow it) before Checkout goes live.

Cloudflare Pages vs Vercel?

Cloudflare wins on unmetered static bandwidth. Vercel wins on Next.js DX and previews. If the site is HTML/CSS/JS, start on Pages. If Cursor just scaffolded Next, start on Vercel and read the Hobby rule before you sell anything.

Is Cloudflare Pages really unlimited?

Static asset requests are. Builds are 500/month. Functions share Workers Free at 100,000 requests/day. Do not promise "unlimited API" on the free plan.

Do I need a credit card?

Usually not for these five free Git hosts. Render may ask for a card if you exceed included bandwidth or pipeline minutes; without one, extra usage can suspend Free services instead of billing.

Originally published on Vibe Coder's Life. I send a short Tuesday/Friday digest on AI coding tools and shipping with agents — subscribe.

Sources: Vercel Hobby / fair-use; Cloudflare Pages limits; Netlify credits; GitHub Pages limits; Render static. Numbers as of 17 Aug 2026 — verify in-product.

Disclosure: Replit and Hostinger links are referral links (same as on our Tools page). Vercel, Netlify, Cloudflare Pages, GitHub Pages, and Render are plain product URLs. Neon signup uses our existing referral (database and storage comes next). No paid placement by the hosting providers. Drafting/editing used AI assistance; the hosting map is from my own shipping work.

Free vibe coding courses (2026)

Viktor Kondas — Fri, 02 Oct 2026 11:00:00 +0000

A practical map of free and low-cost AI coding training — Codecademy, Udemy picks, Anthropic Skilljar, OpenAI Academy, Cursor Workshops, Grow with Google, plus a clear start-here path.

Free vibe coding stack — **Part 2* (courses).*

Part 1 of Free vibe coding stack covered free cloud AI API keys. Keys get you unblocked. Training gets you productive. Part 2 is courses — free vendor workshops first, then structured paths, then paid depth when a sale actually helps.

Start-here (verdict first): Day 0 → Cursor Workshops. Week 1 → Anthropic Skilljar or OpenAI Academy. Then one structured path (Codecademy). Buy Udemy only on a sale for longer video curricula. Grow with Google if you want broader digital / AI literacy — not day-one agent shipping.

Start-here path

Day 0 — Cursor Workshops. Learn the IDE/agent you already build in. Pick Cursor 101, Rules / Skills / MCP, or whatever's live this week.
Week 1 — vendor-native skills. Anthropic Skilljar (Claude Code, MCP, agent skills) or OpenAI Academy — match the model stack you actually use.
Then — structured path. Codecademy Pro paths / agent courses, or the live Building Agentic AI Applications for Beginners bootcamp when a cohort is open (code EARLY20 for 20% off — time-limited; verify at checkout).
Depth / paid sales — Udemy. Prefer instructors Ed Donner and Maximilian Schwarzmüller over random "AI agent masterclass" spam.
Career / fundamentals — Grow with Google (Europe). Broader literacy tracks, not a substitute for Cursor + one vendor academy.

Codecademy

Summary: Interactive lessons, career paths, and occasional live bootcamps. Good when you want a syllabus and quizzes instead of YouTube chaos.

How to register: Create an account via the Codecademy referral link (disclosed). Free catalog exists; Pro unlocks most guided paths. Bootcamps are separate paid cohorts.

Best picks for vibe / agents:

Building Agentic AI Applications for Beginners — 12-week live bootcamp (Python → LLMs / RAG / orchestration). Landing pages have advertised EARLY20 for 20% off; cohort calendars change — if overview says "no upcoming," wait for the next start date.
Self-paced: Learn How to Build AI Agents and the Build AI Agents with OpenAI and LangChain path when you want structured practice without live sessions.

Free vs paid: Free intro lessons; Pro / bootcamp are paid. Don't assume the whole agent track is forever free.

Who it fits: Beginners who want accountability and a curriculum; people who already like Codecademy's UI.

Try Codecademy →

Udemy

Summary: Huge marketplace. Quality varies wildly — filter by instructor, recent updates, and project load, not star spam.

How to register: Browse via the Udemy referral link. Wait for sitewide sales if you can; list prices are theatre.

Instructors / courses verified at write time:

Ed Donner — AI Engineer Core Track: LLM Engineering, RAG, QLoRA, Agents and AI Engineer Agentic Track: The Complete Agent & MCP Course (CrewAI, LangGraph, MCP, multi-framework builds).
Maximilian Schwarzmüller — AI Agents & Workflows: The Practical Guide (clearer conceptual entry if Donner's agentic track feels heavy).

Free vs paid: Paid video (often heavily discounted). Not a free academy.

Who it fits: Self-paced learners who want 10–40 hours of depth after free vendor workshops.

Browse Udemy →

Anthropic Skilljar

Summary: Anthropic's official course hub on Skilljar — Claude for work, Claude Code, Claude API, MCP, AI Fluency, agent skills / subagents.

How to register: Sign up at anthropic.skilljar.com (Skilljar account; Anthropic Console account is separate if you later use the API).

Best courses for vibe coders: Claude Code 101 / Claude Code in Action, Introduction to Model Context Protocol, Introduction to agent skills, Building with the Claude API.

Free vs paid: Learning content is free to access via Skilljar; certificates / progress tracking follow Skilljar's flow.

Who it fits: Anyone shipping with Claude / Claude Code / MCP day to day.

Open Anthropic Skilljar →

OpenAI Academy

Summary: OpenAI's free learning hub for builders and teams using ChatGPT / the OpenAI API / agent patterns.

How to register: Start at academy.openai.com and pick tracks that match your stack (API, assistants / agents, product use).

Free vs paid: Academy content is free; API usage is separate (see Part 1 for free-key options if you're experimenting).

Who it fits: Codex / ChatGPT / OpenAI API users who want vendor-native guidance before buying a third-party bootcamp.

Open OpenAI Academy →

Cursor Workshops

Summary: Free live and recorded sessions from the Cursor team — 101/201, Rules / Skills / MCP, model selection, cloud agents, role-based sessions.

How to register: Browse upcoming events at cursor.com/workshops and join what matches your role.

Best first picks: Cursor 101, Cursor Rules / Skills / MCPs, Model Selection & Token Efficiency.

Free vs paid: Workshops are free; Cursor product plans are separate.

Who it fits: Anyone whose daily driver is Cursor — start here before paying for agent theory elsewhere.

View Cursor Workshops →

Grow with Google (Europe)

Summary: Google's Europe-facing digital skills / AI literacy programs — career and fundamentals oriented more than "ship an agent tonight."

How to register: Start at grow.google/intl/europe/ and pick country / program availability.

Free vs paid: Core Grow tracks are free; partner certificates may vary by country.

Who it fits: Career switchers and teams needing broad AI literacy — not a replacement for Cursor + Skilljar / Academy.

Explore Grow with Google →

Also useful (curated extras)

DeepLearning.AI short courses — prompt engineering / agentic AI videos are free to audit; graded work and certificates often need Pro. Start at deeplearning.ai/courses.
Hugging Face AI Agents Course — free end-to-end agents path; free cert if you complete the stated assignments. huggingface.co/learn/agents-course.
Microsoft Learn / GitHub Skills + AI Agents for Beginners — free Copilot / agent lesson tracks. AI Agents for Beginners.
freeCodeCamp — AI-assisted coding / ML YouTube + curriculum, free with no course upsell. freecodecamp.org.

Comparison

Platform	Free?	Certificate?	Best for	Time to first project
Cursor Workshops	Yes	No (session)	Tool fluency	Same day
Anthropic Skilljar	Yes	Via Skilljar	Claude / MCP / Code	1–3 days
OpenAI Academy	Yes	Varies by track	OpenAI / Codex stack	1–3 days
Codecademy	Partial	Pro / bootcamp	Structured beginners	~1 week
Udemy	No (sales)	Course cert	Deep video curricula	1–2 weeks
Grow with Google	Yes*	Often	Career / literacy	Weeks
Hugging Face Agents	Yes	Free if completed	Open-source agents	~1 week

Checked Aug 2026. Offers and cohorts change — re-verify EARLY20, bootcamp dates, and Academy tracks before you pay.

Summary

Start free where you already work: Cursor Workshops, then Anthropic Skilljar or OpenAI Academy. Add one structured path (Codecademy) if you need a syllabus. Buy Udemy on a sale for Donner / Schwarzmüller depth — don't stack three bootcamps. Part 1's free API keys still matter: training without a key burns paid credits for no reason.

Originally published on Vibe Coder's Life. I send a short Tuesday/Friday digest on AI coding tools and shipping with agents — subscribe.

Disclosure: Codecademy and Udemy links are referral links (same as on our Tools page). Cursor Workshops, Anthropic Skilljar, OpenAI Academy, and Grow with Google are plain product URLs. No paid placement by the course providers for this article. Drafting/editing used AI assistance; the course map is from my own shipping work.