Your first api-turbo request, step by step

Check the prerequisites

The website, catalog and guides are public. API calls require an account and authorization. If sign-up is not offered, request access from the operator. Seeing a model in the catalog does not mean you can call it.

Create a separate development key and check its model restrictions and quota. Use the exact model ID confirmed in your dashboard. Family names, product names and API IDs may differ.

Configure three server-side values

The Base URL is an SDK prefix. For direct HTTP requests, append the endpoint path without duplicating /v1. Keep the key on your server, out of browser bundles, public repositories and shared screenshots.

export API_TURBO_KEY="YOUR_API_KEY"
export API_TURBO_MODEL="YOUR_ENABLED_MODEL_ID"

Send a minimal text request

This example is for an enabled text model that supports Chat Completions. Run it in your terminal or backend after replacing both environment variables. This page never executes requests.

curl --fail-with-body https://api-turbo.com/v1/chat/completions \
  -H "Authorization: Bearer $API_TURBO_KEY" \
  -H "Content-Type: application/json" \
  -d "$(python3 -c 'import json, os; print(json.dumps({"model": os.environ["API_TURBO_MODEL"], "messages": [{"role": "user", "content": "Reply with hello."}]}))')"

Verify the result

Check the HTTP status and response body, then confirm the assistant output. For a conventional text response on this interface, content is usually in choices[0].message.content. Tool calls and other interfaces can return different structures.

After success, check the matching request record and usage in the dashboard. Add context, streaming or tool calls one at a time so failures are easier to isolate.

Next steps

1. 401: verify that you sent an API key, not your account password.
2. Model error: check the exact ID, key restrictions and channel configuration.
3. Open the developer documentation for Python / JavaScript examples and streaming.