Expand description
🌐 LLM Proxy - Unified interface for multiple LLM providers
This module provides a unified proxy for calling various LLMs, including OpenAI, Anthropic, Google Gemini, and local Candle-based models.
“Why talk to one AI when you can talk to them all?” - The Cheet 😺
Re-exports§
pub use claude as anthropic;
Modules§
- candle
- 🤖 Candle Local LLM Provider Implementation
- claude
- Claude API Client - Comprehensive Anthropic Claude integration via raw reqwest
- 🤖 Google Gemini Provider Implementation
- grok
- 🤖 Grok Provider Implementation (X.AI)
- memory
- 🧠 Memory Proxy - Scoped conversation history for LLMs
- oauth
- OAuth 2.0 + PKCE framework for proxy providers.
- ollama
- 🦙 Ollama & LM Studio Provider - Local LLM Auto-Detection
- openai
- 🤖 OpenAI Provider Implementation
- openai_
compat - 🔄 OpenAI-Compatible Request/Response Types
- openrouter
- 🌐 OpenRouter Provider Implementation
- server
- OpenAI-Compatible Proxy Server
- token_
store - Secure token storage for proxy OAuth providers.
- zai
- Z.AI Provider Implementation (Zhipu / GLM family)
Structs§
- LlmMessage
- 💬 A single message in a conversation
- LlmProxy
- 🛠️ Factory for creating LLM providers
- LlmRequest
- 📝 Request structure for LLM completion
- LlmResponse
- 📦 Response structure from LLM completion
- LlmUsage
- 📊 Token usage statistics
Enums§
- LlmRole
- 🎭 Roles in a conversation
Traits§
- LlmProvider
- 🤖 Common interface for all LLM providers