Build against one stable API.

Set the base URL, pass a bearer key, and choose a documented model ID.

curl https://api.aiforpros.dev/v1/chat/completions \
  -H "Authorization: Bearer sk-aifp-YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"top-tools-ai","messages":[{"role":"user","content":"Explain SSE in one paragraph."}]}'

Send your API key in the Authorization header. Keys use the sk-aifp- prefix and are shown once when created.

Authorization: Bearer sk-aifp-YOUR_API_KEY

Call GET /v1/models with your key to receive the models available to your account. The public catalog lists all 22 model IDs.

Send messages to POST /v1/chat/completions. Messages can include system, user, assistant, and tool roles.

Set stream to true. The response uses text/event-stream and ends with the provider completion marker.

curl https://api.aiforpros.dev/v1/chat/completions \
  -H "Authorization: Bearer sk-aifp-YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"top-tools-ai","stream":true,"messages":[{"role":"user","content":"Write one sentence."}]}'

Pass OpenAI-compatible function definitions in the tools array. Tool arguments arrive in the assistant response or streamed deltas.

{
  "model": "top-tools-ai",
  "messages": [{ "role": "user", "content": "Weather in Paris" }],
  "tools": [{
    "type": "function",
    "function": { "name": "get_weather", "parameters": { "type": "object", "properties": { "city": { "type": "string" } } } }
  }]
}

For models marked vision, include image URL content parts beside text content.

{
  "model": "top-tools-ai",
  "messages": [{ "role": "user", "content": [
    { "type": "text", "text": "Describe this image." },
    { "type": "image_url", "image_url": { "url": "https://example.com/image.jpg" } }
  ] }]
}

Use the image generations endpoint with a model that includes image generation access.

curl https://api.aiforpros.dev/v1/images/generations \
  -H "Authorization: Bearer sk-aifp-YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"top-tools-ai","prompt":"A precise technical cutaway drawing"}'

Standard accounts allow 45 requests per minute and 5 concurrent requests. API Limits Plus allows 75 requests per minute and 15 concurrent requests. A sixth concurrent request on a standard account returns 429.

Error responses use stable codes: invalid keys return 401, access denial returns 403, request limits return 429, provider failures return 502, and internal gateway failures return 500.