One key.
One request.

Create an API key.

Generate a token in the console, then keep it in an environment variable. Never expose a secret key in browser code or commit it to source control.

Choose a chat format.

JustAPI supports three chat formats used by common SDK integrations. A model’s detail panel shows which formats it accepts.

{
  "model": "gpt-5.4-mini",
  "messages": [
    { "role": "system", "content": "Be concise." },
    { "role": "user", "content": "Hello" }
  ]
}

{
  "model": "gpt-5.4-mini",
  "instructions": "Be concise.",
  "input": "Hello"
}

{
  "model": "claude-sonnet-5",
  "max_tokens": 1024,
  "system": "Be concise.",
  "messages": [
    { "role": "user", "content": "Hello" }
  ]
}

Chat Completions offers the broadest model compatibility. Check Models for per-model endpoint support.

Responses API provides unified input and built-in conversation state. Check Models for per-model endpoint support.

Anthropic Messages lets an existing Anthropic SDK integration keep its familiar request shape. Check Models for per-model endpoint support.

Send your first request.

This example uses OpenAI Chat Completions. Set https://api.justapi.ai/v1 as the base URL and choose a compatible model ID from the live catalogue.

Read the response.

The response shape follows the protocol selected above. Every format includes usage information for token accounting.

Choose what to ship.

Compare input and output rates, inspect tiered pricing, or connect an agent to the token-efficient JustAPI remote MCP server.