Developer Library
Less guesswork. Better AI products.
Practical articles for your next integration, a clearer API budget and workflows you can trust.
Endpoints and request examples.
Tokens, requests and credit.
Current prices and available routes.
From the field
17 articles
- cost · 7 min
AI API cost calculator: plan tokens, requests and credit
Turn your request volume into a practical credit budget. A worked example, an editable calculator and a checklist for costs that a headline token price misses.
08.09.2026Read article - integration · 7 min
Move your app to a compatible AI API: a migration checklist
A working base URL is the beginning. Verify the request contract, key access, response shape and spending before sending your customers’ traffic.
08.09.2026Read article - troubleshooting · 7 min
API key not working? Diagnose model access and routing
An API key can authenticate while a model request still fails. Identify the failing layer, run the right check and know when a route needs operator attention.
08.09.2026Read article - policy · 8 min
Claude subscription or API key: what the terms actually say
The secondary sources disagree with the primary one. The difference matters if you are choosing how to pay for model access.
05.09.2026Read article - buying · 10 min
Is a cheap Claude API legitimate? Six checks before you send money
Price alone tells you nothing about which one you are looking at. These are checks you can run yourself — including against us, and we fail one of them.
03.09.2026Read article - troubleshooting · 8 min
The ANTHROPIC_BASE_URL 404, and the one path segment behind it
One key, one host, two different base URL strings. The wrong one answers 404, and most clients report that as a connection problem rather than a wrong path.
01.09.2026Read article - monitoring · 13 min
When your uptime probe becomes the outage
Our status page reported most of a twenty-three model catalogue as unstable while customers were getting answers. The probe was measuring itself, three times over.
31.08.2026Read article - reliability · 8 min
What uptime means when your client retries
Same models, same afternoon: 75% of first attempts got an answer, 93% within three. Publishing only one of those is misleading in one direction or the other.
31.08.2026Read article - measurement · 9 min
A p95 from six samples is just your slowest call
A dashboard reading `p95: 35s` from six measurements is not describing a tail. It is naming your worst call something stronger than it is.
31.08.2026Read article - relays · 16 min
AI API relays, explained: how they work and how to audit one
A relay forwards your request to the model provider and returns the answer. That is the whole idea — and every meaningful difference between relays lives in what happens in between.
30.08.2026Read article - testing · 11 min
Canary-testing an AI API relay: a method you can run yourself
You cannot read a relay’s routing table. You can send it a prompt whose answer only one model gives, and see what comes back.
30.08.2026Read article - routing · 12 min
Model authenticity: what happens when one model name has four routes
One model name can sit in front of several upstream channels. Which one answers is a policy decision, and the policy is the thing worth reading.
30.08.2026Read article - pricing · 10 min
We audited our own list prices against the providers, and two were wrong
The number a discount is measured against is supplied by the party claiming the discount. We read ours against the source and published what we found.
30.08.2026Read article - pricing · 9 min
Per 1K or per 1M: the 1000x error waiting in every price comparison
Two prices, two units, one table. The mistake does not look like a mistake — it looks like a spectacular deal.
30.08.2026Read article - pricing · 11 min
Reading a provider’s list price correctly: the conditions nobody quotes
A single number on a pricing page is usually several numbers wearing a coat.
30.08.2026Read article - pricing · 9 min
Prepaid credit or a monthly bill: which billing model fits your workload
The billing model is not an accounting detail. It decides what happens on the day something goes wrong.
30.08.2026Read article - api · 12 min
What “OpenAI-compatible” actually has to mean
Compatible with what, exactly? The request body is the easy half.
30.08.2026Read article