Skip to main content
The DeepInfra API is available at https://api.deepinfra.com. It offers two sets of endpoints:
  1. OpenAI-compatible endpoints at /v1/openai — drop-in replacement for OpenAI chat completions, embeddings, and image generation. Use these with any OpenAI SDK.
  2. DeepInfra native endpoints at /v1/inference/{model_name} — direct access to all model types including speech recognition, object detection, image classification, and more.

​
Authentication

All endpoints require a Bearer token in the Authorization header: 
Authorization: Bearer $DEEPINFRA_TOKEN
Get your API key from the Dashboard.

​
Interactive reference

Use the tabs above to browse and try all available endpoints, generated from our live OpenAPI specification.