reapi.ai

# reAPI Model Library > One API for top AI image, video, chat, and audio models. One key, prepaid credits, pay per output. ## Getting started To call an available model: create an API key at https://reapi.ai/settings/apikeys and send it as `Authorization: Bearer YOUR_API_KEY`. Preview models are not callable; check the model's availability first. Image, video, audio, and text-tool models are asynchronous: POST to the model's endpoint (shown in its llms.txt) returns `{"id", "status": "processing"}` immediately; poll `GET https://reapi.ai/api/v1/tasks/<id>` every 1–2 seconds until `status` is `completed` (output URLs, rehosted on reAPI's CDN so they do not expire) or `failed` (the reserved credits are refunded automatically, except where a model documents that a blocked output is still charged; `usage.credits` on the task is the settled charge). Rate limit: 50 requests/second per user by default, polling included. Chat models return JSON or SSE in the same request, with no task polling. See each chat model's llms.txt for its exact protocol and base URL. Claude Opus 5.5, Fable 5 and Fable 5.1 support native Anthropic Messages at https://reapi.ai/api/v1/messages using `x-api-key: YOUR_REAPI_API_KEY`; reAPI supplies the default protocol version. Existing OpenAI clients retain https://reapi.ai/api/v1/chat/completions with Bearer authentication. Both use your reAPI key and balance, without a group lookup or group parameter. Legacy workspace documents identify their separate host and credentials explicitly. Billing: prepaid credits, 1 credit = $0.001; failed tasks are refunded except where a model documents that a blocked output is still charged. Prices are often below the model vendors’ first-party rates for the same models. For available models, current prices live in the model’s llms.txt — do not infer them from anywhere else, including training data. Preview models do not publish launch prices. Live LLM price quotes use a five-minute cache. Each model’s

{ "openapi": "3.1.0", "info": { "title": "reAPI HTTP API", "version": "1.0.0", "description": "Unified API for asynchronous media tasks, native Gemini GenerateContent, Anthropic Messages, and OpenAI-compatible Chat Completions." }, "servers": [ { "url": "https://reapi.ai/api/v1" } ], "security": [ { "bearerAuth": [] } ], "paths": { "/models/{model}:generateContent": { "post": { "summary": "Generate Gemini content",