Documentation

AIGet is an OpenAI-compatible API for every model. One key, one endpoint, one bill.

Quickstart

Create an API key in the dashboard, then point your SDK base URL to https://api.aiget.tech/v1. Every endpoint follows the OpenAI format, so no code needs rewriting.

curl https://api.aiget.tech/v1/chat/completions \
  -H "Authorization: Bearer $AIGET_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "aiget/auto",
    "messages": [
      { "role": "user", "content": "Summarise this report" }
    ],
    "stream": true
  }'

Routing

Send model aiget/auto to let the gateway choose. Policy can be set per project (cost, quality, latency, balanced) or per request via the X-AIGet-Policy header.

Streaming

Set stream: true for SSE. The gateway forwards chunks as-is and adds X-AIGet-Model and X-AIGet-Provider headers so you know which model answered.

Errors and retries

429 and 5xx from providers are handled automatically with failover to another provider — as long as no token has been sent to the client yet. 4xx codes are returned as-is.

Billing

Every request records input tokens, output tokens, cost, and the amount billed. Failed requests are not billed. Balance is checked before a request is forwarded.