SingleAPI gives your application access to enabled text models through one API key. Start with your first request, using the OpenAI Python SDK or HTTP.
Your first integration
- Create an account, then create a key under API Keys. Copy its secret once and keep it in your application's environment.
- Choose an enabled model under Models. For managed requests, add prepaid credits; alternatively, select a stored BYOK credential.
- Follow the quickstart, then open Usage to inspect requests, tokens and customer charges.
The guides are public. Account, key, model-price and usage pages require sign-in. Examples use placeholders for secrets; never put a real API key in public code or a browser application.
Get started
- How SingleAPI works — the account, model, key and billing flow.
- Make your first request — choose a model, create a key and get a response.
Manage your account
- Choose a model — find a model ID, check prices and understand access.
- Manage API keys — model access, budgets, rotation and revocation.
- Credits and usage — add credits, review requests and download a statement.
- Prompt logs and export — filter captured logs and download the filtered rows as CSV.
- Bring your own key — use your provider account for selected requests.
Use model features
- Stream a response — show text as it arrives.
- Get a JSON response — JSON object and JSON Schema modes.
- Use a function tool — let your application run a function requested by a model.
- Use provider web search — let an enabled provider search the web during a request.
- Fetch a web page — read a specific URL with an enabled Anthropic model.
- Troubleshoot requests — common errors and what to check.
Reference
- API reference — exact URLs, fields, response shapes, limits and errors.
These guides cover SingleAPI's supported OpenAI Chat Completions subset. OpenAI SDK compatibility does not imply support for every OpenAI endpoint or parameter; check the supported feature list.