DEV Community: daniel

Two Months Running an AI-API Reference Site Solo: Google Punished the Launch, Bing Didn't Care

daniel — Fri, 09 Oct 2026 05:55:36 +0000

Two Months Running an AI-API Reference Site Solo: Google Punished the Launch, Bing Didn't Care

I run ai-info — an independent reference site for AI API decisions: pricing breakdowns, model-vs-model comparisons, and guides, aimed at developers who search things like "cheapest coding agent plan" or "is X worth it." No affiliate links (frozen on purpose), nothing for sale in this post. Everything below is a snapshot as of October 8–9, 2026; index states and rankings drift daily, so treat the dates as part of the data.

What shipped

Two months, 42 published pieces (independent slugs on the sitemap as of Oct 8), six sections covering DeepSeek, model comparisons, coding agents, local models, and news. The first batch of 19 posts went out in one burst — deliberately, to give crawlers something to chew on.

The Google arc, in three acts

Act one, the honeymoon. The site got indexed fast — exactly one page: the DeepSeek hub. Not the homepage, not any article. One navigation page. It even earned a brief stretch of impressions. Encouraging, in a narrow way.

Act two, the fade. Sometime around September 18, impressions went to zero and the hub dropped out of the index. No manual action, no message in Search Console, just silence.

Act three, the present. As of the October 8 GSC reading: 3 pages indexed — homepage, news hub, local hub. Zero article pages. Forty-two pieces of content, and Google trusts exactly the skeleton.

Whatever the right label is (scaled-content suspicion, new-domain distrust, or just the post-hangover every search doc warns about), the practical read is the same: this is a site-level verdict, not a per-article one.

What I decided not to do: panic-slow the output to "fix quality." With Bing and direct traffic independently viable and 40+ pieces sitting as a starting library, the play is to keep publishing on a daily floor and point the backlink push at the skeleton pages — homepage and hubs — rather than deep articles. Hubs are what Google trusted once; that's the best signal I have.

The Bing surprise

On September 29 I cross-posted a cost-math piece to dev.to with the canonical pointing back to my original. Bing indexed the copy the same day and counted the backlinks. Google, as expected, never indexed the copy at all — which is exactly how canonical folding is supposed to behave, so no complaint there. It just underlines that the two engines are playing very different games with the same site right now.

That post carried 7 dofollow links back to the site, re-crawled on October 8 and still dofollow. So the channel works as a link source. It does not work as a ranking shortcut — Google never indexed the copy, so none of its on-page text helped my pages. Expect reach ≈ zero, expect credit = the links themselves. Budget your effort accordingly.

The workflow that makes 42 posts survivable

Volume only works if per-post quality has teeth. Three rules doing the heavy lifting:

Narrow verification. Every post gets 3–5 hard numbers checked directly against official sources. Background figures cite the outlet they came from, by name. Verification stamps only go on numbers actually reconciled.
The Chinese-media translation gap. I monitor Chinese tech outlets and vendor announcements. A DeepSeek pricing change announced in Chinese typically gets real English coverage a day or two later. Writing for that 6–48 hour window is the closest thing to an unfair advantage a solo operator gets.
Dated caveats everywhere. Every post carries an "as of" statement. I'm only accountable for correctness at publish time — prices and plans drift, and pretending otherwise is how a reference site loses trust.

Honest scoreboard

Two months in: Google 0 (3 skeleton pages, no impressions), Bing 1 (same-day indexing, backlinks counted), direct and referral traffic small but real and growing. The bet underneath all of it: decision-intent queries around AI APIs are still under-served in English, and backlinks plus time still compound for genuinely useful pages.

Check back in six months. Or don't — I'll be posting the numbers either way.

Snapshot caveats: index counts from GSC on Oct 8–9, 2026; sitemap slug count from a live crawl on Oct 8, 2026; dev.to link attributes re-verified Oct 8, 2026. All of this changes — verify anything you'd cite.

Your DeepSeek V4 Bill Can Swing 2x by the Hour — the Actual Math Across Four APIs

daniel — Tue, 29 Sep 2026 03:45:05 +0000

Full disclosure up front: I run ai-info.fastget.link, where I track AI API pricing and publish the DeepSeek V4 cost calculator this post is based on. All rates below were checked against official pricing pages and live dashboards on 2 Sep 2026 — if you're reading this later, treat them as a snapshot, not gospel. No affiliate links anywhere.

The pricing page is lying to you (sort of)

DeepSeek doesn't have one price for V4. It has four, and which one you pay depends on decisions most developers never consciously make.

Take V4 Flash, the cheap workhorse model. DeepSeek's official list rate is $0.33 input / $0.99 output per million tokens — except that's not what you'll actually pay. DeepSeek runs peak/off-peak billing: off-peak hours cost ×0.67 of list ($0.22/$0.66), peak hours cost ×1.33 ($0.44/$1.32). So the same API call costs $0.22/M input at 3 AM and $0.44/M at 2 PM. Same model, same tokens, 2× apart. (The exact windows and how they map to your timezone are in my off-peak pricing breakdown.)

And that's just DeepSeek's own portal. The same model ID is resold elsewhere:

Provider	Flash (in/out per M)	Pro (in/out per M)
DeepSeek official (list)	$0.33 / $0.99	$0.99 / $2.97
DeepSeek official, off-peak (×0.67)	$0.22 / $0.66	$0.66 / $1.98
DeepSeek official, peak (×1.33)	$0.44 / $1.32	$1.32 / $3.96
OpenRouter (cheapest host)	$0.05 / $0.16	$0.58 / $1.74
Novita	flat rate, see dashboard	flat rate, see dashboard

Yes, you read the OpenRouter row right. The cheapest host on OpenRouter serves deepseek-v4-flash-0731 at $0.05/$0.16 — roughly a quarter of DeepSeek's off-peak rate. For Pro (deepseek-v4-pro-0813), the cheapest route is Alibaba at $0.58/$1.74, which actually undercuts DeepSeek's own off-peak rate.

So which number ends up on your invoice? That depends on three choices:

1. Official vs. reseller is a cache trade, not just a price trade

The per-token table says "reseller wins." But look at cache pricing before you switch: DeepSeek official sells cache hits at $0.007–$0.044/M — dramatically cheaper than anything resellers offer. If your app reuses long prompts (RAG with a fat system prompt, agents with big tool schemas, long-running chat threads), your real bill at DeepSeek official can come in far below the naive per-token estimate, sometimes below the reseller's headline rate.

Rule of thumb I've settled on: high cache-hit workload → official; stateless, prompt-diverse workload → the cheapest reseller route.

2. Peak windows are a scheduling problem, and scheduling is free

DeepSeek's off-peak discount is the cheapest optimization in the entire AI API market, because it costs zero code quality. Batch jobs, eval runs, embeddings backfills — none of them care whether they run at 3 AM. If half your Flash tokens are batch-shaped, moving them off-peak cuts that half's cost by half ($0.22 vs $0.44 per million input — a 2× spread).

The Pro model makes this more dramatic: official Pro ranges from $1.98/M output off-peak to $3.96/M at peak. Your nightly reasoning batch and your midday interactive traffic are not the same product and shouldn't share a price.

3. Watch the model ID, not the marketing name

One trap worth its own section: on OpenRouter, the undated deepseek-v4-pro ID routes from $0.87/$1.74 — above DeepSeek's own off-peak rate — while the dated deepseek-v4-pro-0813 starts at $0.58/$1.74 via Alibaba. Same model, different ID, different routing pool, different floor price. Pick by model ID and check which host actually serves the route; OpenRouter's "cheapest host" changes and per-route pricing isn't uniform.

Running your own numbers

The calculator on my site takes your monthly input/output token volumes and returns totals for both models across providers, with the peak/off-peak midpoint math built in. To give you scale: at a fairly typical solo-project load of 10M input / 5M output tokens per month on Flash, the spread between OpenRouter's cheapest host and DeepSeek's peak-hour official rate is the difference between roughly $1 and $11 a month. On Pro, the same spread is tens of dollars.

One caveat on the table: Krater.ai is subscription-based ($20/mo for 1,500 credits across 350+ models), so it's excluded from per-token comparisons. If your usage is spiky rather than steady, a subscription can beat all the per-token rates above — but that's a different math problem.

The short version

The "DeepSeek V4 price" is a range, not a number: ×0.67 off-peak to ×1.33 peak on official, and resellers undercut even the off-peak rate by 4× on Flash.
Cache-heavy workloads belong on DeepSeek official ($0.007–$0.044/M cache hits); stateless workloads should shop routes.
Schedule batch work into off-peak windows — it's the only API optimization that costs nothing.
On OpenRouter, quote the dated model ID and check the host.

Prices shift frequently on every one of these providers — I re-verify against official pages and dashboards regularly, and the calculator gets updated when they move. If a number here looks stale, the calculator page is the live copy.

Cross-posted from ai-info.fastget.link, my site on AI API pricing and decisions — the canonical version of this post lives there.