DEV Community: Sameer Moin

Stripe Is Taking More Than You Think. Here's the Actual Math.

Sameer Moin — Wed, 26 Aug 2026 22:17:57 +0000

You sent an invoice for $500.

Stripe deposited $485.

You stared at the number, shrugged, and moved on.

Most freelancers do. And that's exactly how Stripe fees quietly eat into your income month after month without you ever sitting down to calculate how much you're actually losing.

Let me break it down properly.

The Standard Rate Is Just the Starting Point

Stripe's base rate in the US is 2.9% + $0.30 per successful transaction.

On a $500 invoice that's:

2.9% of $500 = $14.50
Fixed fee = $0.30
You receive: $485.20

Fine. But this is only true if your client is in the same country, paying in the same currency, with a domestic card.

The moment any of that changes, the fees stack.

How Fees Stack Up

Scenario	Extra Fee	Effective Rate on $500
Standard domestic	2.9% + $0.30	$485.20 net
International card	+1.5%	$477.70 net
Currency conversion	+1.0%	$472.70 net
Both + conversion	+2.5%	$472.70 net

That $500 invoice to an international client paying in a different currency? You might actually receive around $472.

The Refund Problem Nobody Mentions

If your client asks for a refund, Stripe does not return the processing fee.

You refund the full $500 to the client. But the 2.9% + $0.30 you already paid? Gone.

So you lose the money AND the fee. For high-refund-rate work, this is worth building into your pricing.

The Right Way to Invoice

If you want to receive exactly $500, you can't charge $500.

The formula is:

Gross to charge = (Desired net + $0.30) / (1 - 0.029)

To receive $500 net, you need to charge $515.25.

Most freelancers don't know this formula exists. They just charge what they want to earn and silently lose money on every single payment.

A Free Calculator for This

I built a Stripe Fee Calculator as part of ToolsFusion that handles both directions.

Enter what you're charging and see your exact net payout. Or enter what you want to receive and get the exact gross amount to put on the invoice.

It covers:

US, UK, Europe, Australia, Canada
International card fees
Currency conversion fees
Recurring charges
Stripe Invoices

Everything runs in your browser. No account needed.

🔗 thetoolsfusion.com/calculator/stripe-fee-calculator

How to Actually Price an AI Voice Agent for a Client (And Not Lose Money Doing It)

Sameer Moin — Sun, 23 Aug 2026 11:18:04 +0000

You just landed a client who wants an AI voice agent. Nice.

Then they ask what it will cost to run monthly and you realize you have no clean answer.

This is more common than people admit. The platforms have confusing pricing, most guides just quote the headline rate, and there is no standard for what to charge clients on top of that.

Let me break down how this actually works, and share a free tool that makes it easier.

The Problem With Advertised Platform Rates

Every major platform shows a per-minute rate. And almost every one of them is incomplete.

Platform	Advertised	Real all-in estimate
Vapi	$0.05/min	$0.10–$0.30/min
Retell AI	$0.055–$0.07/min	$0.11–$0.31/min
Bland AI	~$0.09/min	Higher at low volume
Synthflow	$0.13–$0.20/min	Closest to actual

The reason there is such a gap: a single call has 4 cost layers that platforms often bill separately.

The 4 Cost Components of Every AI Voice Call

Speech-to-text

Turns the caller's voice into text the model can read. Usually the cheapest part, around $0.003 to $0.02/min.
Language model

Decides what the agent actually says. The most variable cost depending on which model you use.
Text-to-speech

Generates the agent's voice. Varies by provider and voice quality.
Telephony
Actual phone carrier minutes. This is the one most people forget. Even platforms that bundle most things will often bill telephony separately.

When you add all four together, the real cost is almost always higher than what you see on the landing page.

Flat Retainer vs Usage-Based

Both work. But the choice has real consequences.

Usage-based: scales with the client and protects your margin as they grow.

Flat retainer: simpler for the client to budget around. But you need to know your break-even call volume first. That is the point where a flat fee stops covering your actual platform cost. If you do not know that number before you quote, you are just guessing.

A Free Calculator for This

I saw this question come up a lot with no clean tool to answer it, so I built one as part of ToolsFusion.

The AI Voice Agent Pricing Calculator takes:

Platform (Vapi, Retell, Bland, Synthflow, or custom rates)
Expected monthly call volume
Average call length
Your setup hours and hourly rate
Target margin or markup

And gives you:

Recommended per-minute client rate
Flat retainer option
One-time setup fee
Break-even call volume

Everything runs in your browser. No signup, no data sent anywhere.

If you are building or selling AI voice agents, bookmark it. And if you spot anything off in the platform estimates, drop a comment. I keep them updated.

🔗 thetoolsfusion.com/calculator/ai-voice-agent-pricing-calculator

I launched ToolsFusion on Product Hunt today — 16 free browser-side tools for devs and freelancers

Sameer Moin — Tue, 18 Aug 2026 19:55:14 +0000

I've been building ToolsFusion quietly for the past few months alongside my day job as a Laravel developer. Today I finally launched it properly on Product Hunt, and I wanted to share it with the Dev.to community because honestly, this is the crowd I built it for.

What is ToolsFusion?

ToolsFusion is a free collection of 16 tools for developers, freelancers, and small business owners. Things like:

AI Token Cost Calculator - paste your prompt, get the exact cost across GPT-4, Claude, Gemini
Invoice PDF Generator - fill in your details, download a clean PDF. No account.
Stripe & PayPal Fee Calculator - know exactly what you'll receive before sending an invoice
Image Compressor - drag, compress, download. Nothing leaves your browser.
OG Meta Tag Previewer - see how your links look on Twitter, Facebook, LinkedIn before publishing
QR Code Generator
Barcode Generator
Freelance Hourly Rate Calculator
And 8 more

The thing I'm most proud of

Every single tool runs 100% in the browser.

The Invoice PDF is generated on your device. The Image Compressor never uploads your file. The AI Token Counter tokenizes your prompt locally. No data ever touches a server.

This isn't a privacy policy promise - it's how the code is written. I wanted to build tools I'd actually trust myself.

The tech behind it

The site is built with Laravel and Blade on the backend, but the tools themselves are pure JavaScript - no frameworks, no dependencies where possible. Browser APIs do the heavy lifting: the File API for image compression, jsPDF for invoice generation, the canvas element for QR codes.

The goal was always: fast, lightweight, zero friction.

Why I built this

I kept switching between random websites full of ads and paywalls just to calculate a Stripe fee or preview an OG tag. Every tool existed somewhere, but scattered, slow, and either behind a login or drowning in popups.

So I built the version I wanted to exist.

What's next

I'm planning to keep adding tools based on what the community actually needs. If there's something you find yourself Googling repeatedly or using a janky online tool for - I'd genuinely love to know in the comments.

Support the launch

If you find any of the tools useful, I'd really appreciate your support on Product Hunt today - it's launch day and every upvote counts 🙏

🔗 https://thetoolsfusion.com
🚀 https://www.producthunt.com/posts/toolsfusion

Your Links Look Broken When Shared — Here's the 5-Minute Fix

Sameer Moin — Sun, 16 Aug 2026 12:15:01 +0000

The broken card nobody notices until it's too late

You share a link on Twitter, LinkedIn, or in a Slack channel.
Instead of a clean card with a title, description, and image,
people see a bare URL — or worse, a wrong title and a broken
image placeholder.

This isn't a platform bug. It's a missing five lines of HTML
that you control entirely.

What Open Graph tags actually are

When a URL gets shared anywhere on the web, the platform
doesn't display the URL itself — it sends a bot to fetch the
page and looks for specific meta tags in the <head> to know
what to show. These are Open Graph tags, originally built by
Facebook but now read by every major platform.

Rough rules of thumb for what happens when they're missing:

No tags at all — platforms show nothing, or guess badly from the page title
Missing og:image — a random image gets pulled from the page, often a tiny logo that looks terrible at card size
Wrong tag values — the card renders, but with stale or incorrect info

The same nine tags work across nearly every platform. This is
normal and expected — one small block of markup, reused
everywhere.

Why title tags and image tags are handled differently

Most people get the text tags right and the image tag wrong.
This is a mistake because:

og:image is the most visually prominent part of the card and the most fragile
Dimensions must be 1200×630px — anything else gets cropped unpredictably
The URL must be absolute (https://yoursite.com/og.jpg), not relative — bots don't know your domain
Some image hosts block hotlinking, so the card shows broken even though the URL works in a browser

A missing or malformed og:image can undo otherwise correct
title and description tags.

The honest problem with cached previews

Most people assume fixing their tags fixes their card
immediately. Most are wrong.

Once a platform fetches a URL and caches the result, changing
the tags doesn't refresh the card — even after everything is
corrected. Today that means:

Facebook / LinkedIn — cached until you use the Sharing Debugger and click "Scrape Again"
Twitter / X — cached until you run it through the Twitter Card Validator
Slack / WhatsApp — cached until the link is pasted fresh in a new message, and even then it clears on its own timeline

A tool that shows a "live preview" without warning about
caching is giving a preview of the fix, not the reality of
what viewers will see.

What to check before sharing a URL publicly

Nine tags worth knowing before you publish:

<!-- Open Graph — Facebook, LinkedIn, and most platforms -->
<meta property="og:title" content="Your Page Title" />
<meta property="og:description" content="Your description." />
<meta property="og:image" content="https://yoursite.com/og.jpg" />
<meta property="og:url" content="https://yoursite.com/page" />
<meta property="og:type" content="website" />

<!-- Twitter Card — Twitter reads these first -->
<meta name="twitter:card" content="summary_large_image" />
<meta name="twitter:title" content="Your Page Title" />
<meta name="twitter:description" content="Your description." />
<meta name="twitter:image" content="https://yoursite.com/og.jpg" />

Show how these render differently across Twitter, Facebook,
and LinkedIn using the same set of values.

Practical situations where this matters

Sharing a landing page for the first time and wanting it to look intentional
Debugging why a link looks fine on Facebook but broken on Twitter
Migrating a site and needing to confirm nothing regressed
Working in Next.js or Laravel and wanting framework-correct syntax, not just raw HTML
Auditing a whole site's pages before a launch or campaign

Try it yourself

I built a free OG Meta Tag Builder that shows a live preview
on Twitter, Facebook, and LinkedIn simultaneously. It checks
your image's actual pixel dimensions against the 1200×630px
standard and flags issues automatically. Copy the generated
tags in HTML, Next.js, or Laravel Blade format — ready to
paste into your project.

Free OG Meta Tag Generator & Social Preview Tool — ToolsFusion

Build Open Graph and Twitter Card meta tags with a live preview of how your link will look on Twitter, Facebook, and LinkedIn. Free, no signup, export ready-to-use HTML.

thetoolsfusion.com

No signup. Runs entirely in your browser.

A few things worth knowing

Platform	Reads first	Falls back to
Twitter / X	`twitter:*` tags	`og:*` tags
Facebook	`og:*` tags	Page title
LinkedIn	`og:*` tags	Page title
Slack	`og:*` tags	Page title
Discord	`og:*` tags	Page title
WhatsApp	`og:*` tags	Page title

Standards change slowly here, but hotlink policies and crawler
behavior on individual platforms do shift. Check the tool for
the most current preview behavior.

The bottom line

Adding Open Graph tags doesn't just fix how a link looks. It
changes how people decide whether to click at all — a card
with a real title, description, and image reads as
intentional, which makes you look like someone who ships
finished things.

AI API Costs Are Surprising People — Here's How to Check First

Sameer Moin — Fri, 14 Aug 2026 19:34:01 +0000

The bill that surprises everyone eventually

You're building something with the OpenAI or Anthropic API.
It's working well. Then you check your usage dashboard and
the number is... not what you expected.

This happens to almost everyone who builds with AI APIs for
the first time. Not because the pricing is hidden - it's
published clearly - but because tokens are not an intuitive
unit of measurement, and the costs compound in ways that
aren't obvious until you're already past them.

What a token actually is

Tokens aren't words. They aren't characters. They're chunks
of text a model's tokenizer splits your input into, and the
exact split depends on the model.

Rough rules of thumb:

~4 characters per token in English prose
~0.75 words per token
Code tokenizes differently - often more tokens per line than equivalent prose
Non-English text can be significantly more tokens per word
Whitespace, punctuation, and formatting all count

The same sentence can be a different number of tokens
depending on which model you're using. This is normal and
expected.

Why input and output tokens are priced differently

Most people focus on input cost (your prompt) and forget
about output cost (the completion). This is a mistake because:

Output tokens are almost always more expensive than input
GPT-4o: input $2.50/1M, output $10.00/1M - 4x more expensive
Claude Opus 4: input $15/1M, output $75/1M - 5x more expensive
A long completion can cost more than the prompt that triggered it

The honest problem with token counters

Most AI token counters claim to give you an exact count for
any model. Most are lying, or at best, guessing.

Exact token counts are only possible when you have access to
the model's actual tokenizer. Today that means:

OpenAI - exact, because they publish their tokenizer (tiktoken) and it can run client-side
Anthropic (Claude) - estimated only, because they don't publish a portable client-side tokenizer
Google (Gemini) - estimated only, same reason

A tool that claims to give you an exact Claude token count
is giving you a confident approximation. That's fine as long
as it's honest about it. Most aren't.

What to check before a large API call

Three things worth knowing before you send:

How many tokens is this prompt?
How much of the model's context window does that use?
What will this cost if the completion is X tokens long?

Show how each of these plays out differently across GPT-4o,
Claude Sonnet 4, and Gemini 2.5 Flash using the same prompt.

Practical situations where this matters

Sending a large codebase for review - will it fit in context?
Building a RAG system - how much does each retrieval cost at scale?
Comparing providers before committing to one for a high-volume task
Debugging a prompt that keeps getting truncated
Estimating monthly API costs before building a feature

Try it yourself

"I built a free token counter that handles all three major
providers, is upfront about which counts are exact vs.
estimated, and lets you add custom pricing if the built-in
table is out of date. No signup, runs entirely in your browser."

Free AI Token Counter & API Cost Calculator — ToolsFusion

Count tokens for GPT, Claude, and Gemini prompts and estimate API costs before you call. Free, no signup, runs entirely in your browser.

thetoolsfusion.com

A few numbers worth knowing

Model	Context Window	Input (per 1M)	Output (per 1M)
GPT-4o	128K tokens	$2.50	$10.00
GPT-4o mini	128K tokens	$0.15	$0.60
Claude Opus 4	200K tokens	$15.00	$75.00
Claude Sonnet 4	200K tokens	$3.00	$15.00
Gemini 2.5 Pro	1M tokens	$1.25	$10.00
Gemini 2.5 Flash	1M tokens	$0.075	$0.30

Prices change. Check the calculator for the most current
verified rates.

The bottom line

Knowing your token count before you send doesn't just save
money on individual calls. It changes how you think about
prompt design - you start making deliberate tradeoffs between
context richness and cost, which makes you a better AI
application builder.

The Freelance Rate Mistake That's Costing You Thousands

Sameer Moin — Thu, 13 Aug 2026 19:55:25 +0000

The math every freelancer gets wrong

Most people start freelancing by doing this:

Target income ÷ 2,000 hours = hourly rate

It seems logical. It's wrong. And it's costing you thousands
of dollars a year without you realising it.

The 2,000-hour formula assumes:

You are billable 100% of the time
You have zero business expenses
Someone else pays for your health insurance, equipment, and software
You get paid vacation and sick days

None of that is true when you freelance.

What 2,000 hours actually looks like

Walk through a real example:

40 hours/week × 50 weeks = 2,000 hours
But how many of those are actually billable?
- Writing proposals: ~3 hrs/week
- Invoicing and chasing payments: ~2 hrs/week
- Client emails and admin: ~3 hrs/week
- Marketing and finding new clients: ~3 hrs/week
- Learning and keeping skills current: ~2 hrs/week
That's 13 hours/week gone before you do a minute of paid work
Real billable hours: ~27/week = ~1,350/year
Not 2,000

The five variables your rate actually depends on

Target annual net income (what lands in your bank account after everything)
Annual business overhead (software, hardware, insurance, accountant fees)
Tax reserve (25–30% in most countries — self-employment tax hits hard)
Time off (vacations, public holidays, sick days — all unpaid now)
Billable ratio (the percentage of your working hours you can actually charge for — realistically 60–75%)

The formula that actually works

Full formula step by step:

Gross Revenue Target =
(Target Net Income + Annual Overhead) ÷ (1 − Tax Rate)

Total Billable Hours =
(52 − Weeks Off) × Hours Per Week × Billable Ratio

Minimum Hourly Rate =
Gross Revenue Target ÷ Total Billable Hours

A real example with real numbers

Walk through a concrete calculation:

Target net income: $60,000/year
Annual overhead: $8,000
Tax reserve: 28%
Weeks off: 4
Hours per week: 40
Billable ratio: 65%

The answer is NOT $30/hr (what 2,000 hours gives you)
— it's closer to $58/hr. Nearly double.

What happens when you undercharge

The compounding effect:

$30/hr × 1,350 billable hours = $40,500 gross
After taxes (~28%): $29,160 net
After overhead ($8,000): $21,160 take-home
You targeted $60,000. You got $21,160.

That's not a small rounding error. That's a life-altering gap.

Try it yourself

Introduce the calculator naturally:
"I built a free calculator that does all of this automatically —
plug in your numbers and it tells you your actual minimum rate.
No signup, runs in your browser."

Freelance Hourly Rate Calculator — ToolsFusion

Calculate your minimum freelance hourly rate based on income goals, overhead, taxes, time off, and billable ratio — entirely in your browser. Free calculator, no account required.

thetoolsfusion.com

One more thing: your minimum rate is a floor, not a ceiling

Your calculated rate is the minimum you can charge and still
hit your income target. It says nothing about what the market
will pay, what your experience is worth, or what premium
specialist skills command.

Most freelancers who do this exercise for the first time
discover they've been undercharging. Some by a little. Some
by a lot.

The number doesn't lie. Your gut feeling about "what feels
reasonable to charge" usually does.

DEV Community: Sameer Moin

Stripe Is Taking More Than You Think. Here's the Actual Math.

The Standard Rate Is Just the Starting Point

How Fees Stack Up

The Refund Problem Nobody Mentions

The Right Way to Invoice

A Free Calculator for This

How to Actually Price an AI Voice Agent for a Client (And Not Lose Money Doing It)

The Problem With Advertised Platform Rates

The 4 Cost Components of Every AI Voice Call

Speech-to-text

Language model

Text-to-speech

Flat Retainer vs Usage-Based

A Free Calculator for This

I launched ToolsFusion on Product Hunt today — 16 free browser-side tools for devs and freelancers

What is ToolsFusion?

The thing I'm most proud of

The tech behind it

Why I built this

What's next

Support the launch

Your Links Look Broken When Shared — Here's the 5-Minute Fix

The broken card nobody notices until it's too late

What Open Graph tags actually are

Why title tags and image tags are handled differently

The honest problem with cached previews

What to check before sharing a URL publicly

Practical situations where this matters

Try it yourself

Free OG Meta Tag Generator & Social Preview Tool — ToolsFusion

A few things worth knowing

The bottom line

AI API Costs Are Surprising People — Here's How to Check First

The bill that surprises everyone eventually

What a token actually is

Why input and output tokens are priced differently

The honest problem with token counters

What to check before a large API call

Practical situations where this matters

Try it yourself

Free AI Token Counter & API Cost Calculator — ToolsFusion

A few numbers worth knowing

The bottom line

The Freelance Rate Mistake That's Costing You Thousands

The math every freelancer gets wrong

What 2,000 hours actually looks like

The five variables your rate actually depends on

The formula that actually works

A real example with real numbers

What happens when you undercharge

Try it yourself

Freelance Hourly Rate Calculator — ToolsFusion

One more thing: your minimum rate is a floor, not a ceiling