APIrove

Home / API documentation

DEVELOPER DOCUMENTATION

API documentation

Use an APIrove key to call a model enabled for your account. Confirm the model ID, price, permissions, and supported endpoint before integrating it.

API quickstart

Sign in, confirm a model, create an API key, and send the first request. The full walkthrough is the first-call guide. The homepage does not keep a second copy of the curl examples.

Two key groups

New keys default to the Text group for GPT text models and can also generate images (Codex defaults to Image 2.6). Choose the Full group for Claude. Existing keys remain valid. The comparison is in the beginner guide and the pro guide.

Guides

Base URL

https://apirove.com/v1

The homepage keeps the Base URL only. Full request examples stay in the OpenAI-compatible API guide. Key groups are in Two key groups. Cache rates, image-size prices, and low-balance notes stay on the pricing page.

Authenticate with an API key

For OpenAI-compatible endpoints, pass the API key as a Bearer token in the request header:

Authorization: Bearer YOUR_API_KEY

Create or revoke keys in API key management. Store the key on your server. Never embed a long-lived key in browser JavaScript or a public mobile client.

Text request endpoints

Verified examples include the Responses API and Chat Completions. A model may support only one format; check the model details before sending a request.

  • POST /v1/responses
  • POST /v1/chat/completions
  • GET /v1/models — requires API key authentication

See the OpenAI-compatible API integration guide for cURL and SDK examples.

Models, access, and pricing

Public USD unit prices are on the models and pricing pages. Your account group controls which models you can call. Sign in to check balance, access, and usage.

Error handling

For 401, check the API key. For 403, verify model and account permissions. For 429, follow the response guidance and review account limits. For 5xx, capture any request ID and avoid blindly retrying non-idempotent requests.

Streaming, parameters, and model features vary by model. Validate each model and endpoint combination before production use.