Documentation

One API key, four wire formats, and a router in front of every model provider. Start from what you are trying to do.

Connect a coding toolPoint Claude Code, Codex, Gemini CLI or any OpenAI-compatible client at VipAI with two environment variables.Read →Call the API directlyAnthropic native, OpenAI Chat Completions, OpenAI Responses and Gemini native formats.Read →Understand your billWhy one message produces several billing rows, how cache reads are priced, and when Fast mode costs more.Read →

Three steps to your first request#

Nothing here needs a paid model. The complimentary lanes are enough to complete all three steps.

Create a key#

Sign in and open the API keys page in the dashboard. Keys are 48-character strings and are shown once — store the key in a password manager or environment variable, never in a repository.

Point one tool at the router#

Set the base URL and the key. If you already run an agent tool, follow the Claude Code walkthrough — it is the shortest path to a working setup.

Send one request#

A 200 with a completion means the whole chain works.

request.sh

curl https://api.vipai.site/v1/chat/completions \
  -H "Authorization: Bearer $VIPAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5.6-sol",
    "messages": [{ "role": "user", "content": "Say hello in one word." }]
  }'

Everything you need to remember#

Base URL: https://api.vipai.site
OpenAI SDK base URL: https://api.vipai.site/v1 — the OpenAI client appends /v1 itself
Anthropic SDK base URL: https://api.vipai.site — no path suffix
Auth header: Anthropic x-api-key · OpenAI and Gemini Authorization: Bearer
Key length: 48 characters
Server timeout: 600 s for completions, 300 s for image generation

Every page in this build#

Page	What it answers	Length
Connect Claude Code	How do I get an agent tool talking to the router?	4 min
API integration	Which endpoint, header and body for each protocol?	9 min
Models & routing	Which model ids exist and which protocol does each need?	5 min
Rate limits	How many requests can I send on the free models?	2 min
Billing questions	Why is this invoice the size it is?	4 min