AI API knowledge center: setup, token billing and testing
From your first request to a research batch, understand the protocol and billing, then validate a small task. Explore original setup, troubleshooting and testing guides.
AI API setup: Base URL, API Key and model ID
Connect an OpenAI-compatible AI API safely. Understand Base URLs, model IDs, protocol selection and a minimal first request.
Token billing explained: input, output and cache
Understand AI API input tokens, output tokens, cache reads and cache writes. Estimate costs without confusing account credit with token usage.
AI API errors: fix 401, 404, 429 and timeouts
Troubleshoot AI API authentication, model not found, rate limits, browser CORS and streaming failures with a practical diagnostic checklist.
Claude API for coding: Messages, compatible routes and setup
A practical Claude API guide for coding workflows: Messages vs compatible routes, model IDs, prompt caching and tool support checks.
Gemini API setup: native vs OpenAI-compatible requests
Connect Gemini API to a compatible client and check route, model ID, image support, streaming and token usage before scaling.
AI APIs for research: batch documents and code safely
Plan a reproducible AI API workflow for research: document extraction, experimental code, bounded concurrency, retries and cost records.
How to test an AI API: connectivity is not model verification
Understand what API connectivity, latency and model fingerprint tests can establish, and how to read a model-consistency claim critically.
Tools and next steps
API connectivity and streaming checker · Models and pricing · Client tutorials
Get an API key ↗