API Overview

💫 Relay Interfaces

• Chat

  ---

  Support for multiple mainstream chat model formats:

  OpenAI Chat → OpenAI Responses → Anthropic Chat → Deepseek Chat → Google Chat →

• Embeddings

  ---

  Text vector embedding services:

  OpenAI Embeddings →

• Rerank

  ---

  Search result reranking services:

  Jina AI Rerank → Cohere Rerank → Xinference Rerank →

• Realtime Chat

  ---

  Support for streaming real-time conversations:

  OpenAI Realtime →

• Image

  ---

  AI image generation services:

  OpenAI Image → Midjourney Proxy →

• Audio

  ---

  Speech-related services:

  OpenAI Audio →

• Music

  ---

  AI music generation services:

  Suno API →

• Video

  ---

  AI video generation and query services:

  Generate Video → Query Video →

🖥️ Frontend Interfaces

• API Reference

  ---

  The system adopts a four-level auth mechanism: public, user, admin, and root. The interface prefix is unified as http(s)://<your-domain> ( HTTPS should be used in production to protect auth tokens. HTTP is recommended only for development. )

  Auth System Description → Available Models List →

---

📖 Interface Description

Interface Types

Lynx Api provides two main types of interfaces:

1. Relay Interfaces: For AI model calls, supporting multiple mainstream model formats
2. Frontend Interfaces: For supporting Web interface functionality calls, providing complete frontend functionality support

Feature Support Indicators

In the API documentation, we use the following icons to indicate feature support status:

• ✅ Supported: This feature is fully implemented and available for use
• 🟡 Partially Supported: Feature is available but has limitations or provides only partial capabilities
• ❌ Not Supported: This feature is under development or planned for development

Quick Start

1. Browse the cards above to select the interface you need to use
2. Click "View Details" on the corresponding card to learn specific usage
3. Follow the documentation instructions for interface calls