Rust LLM API Router：用Rust构建高性能大语言模型统一路由网关

This is a Rust-developed LLM API routing gateway that supports 34 AI service providers. It offers auto failover, intelligent load balancing, and cost optimization features, allowing developers to manage multiple LLM services seamlessly with just one API key. The gateway unifies various providers behind a standardized OpenAI-compatible interface, simplifying integration and maintenance.

With the rapid development of LLM technology, dozens of AI service providers (like OpenAI, Anthropic, Google, Groq) have emerged, each with unique API interfaces, authentication methods, and pricing strategies. Key pain points for developers include:

• Multi-key management increasing configuration complexity
• Manual failover or complex retry logic when a provider is down
• Lack of automatic cost optimization across different models
• Interface differences adding code maintenance costs. Rust LLM API Router addresses these issues by acting as a smart gateway layer.

The gateway's core functions include:

1. Unified Endpoint: A local endpoint (http://localhost:8080/v1/chat/completions) fully compatible with OpenAI's API, enabling seamless migration of existing OpenAI SDK code.
2. Auto Failover: Transparently routes requests to backup providers when a service is down or slow, ensuring availability.
3. Intelligent Load Balancing: Supports strategies like cost priority (choose cheapest), performance priority (fastest response), and quality priority (model capability scoring).

Multiple deployment methods are available:

• Docker (Recommended): One command to run: docker run -d -p 8080:8080 ghcr.io/xavicode1000/rust-llm-api-router:latest
• Binary Install: Use scripts for user-level (curl -sS https://raw.githubusercontent.com/XaviCode1000/Rust-LLM-Api-Router/main/scripts/install.sh | sh) or system-level (sh -s -- --bin-dir /usr/local/bin) installation.
• Source Compilation: Clone the repo (git clone https://github.com/XaviCode1000/Rust-LLM-Api-Router.git) and build with cargo build --release (requires Rust toolchain).

Two configuration modes are supported:

• Interactive (For Newbies): Use commands like llm-router provider add --interactive or llm-router account add --interactive to complete setup via prompts.
• Command-Line (For Automation): Directly configure via parameters, e.g., llm-router provider add --id groq --name "Groq" --url "https://api.groq.com/openai/v1" or llm-router account add --id mi-key --provider groq --api-key $GROQ_API_KEY.

Rust was chosen for its advantages:

• Memory Safety: Compile-time checks eliminate memory leaks and data races, ensuring long-term stability.
• High Performance: Compiled binary runs efficiently, handling high concurrency with low latency.
• Lightweight: Single binary with no runtime dependencies, small Docker image size.
• Cross-Platform: Supports Linux, macOS, Windows via cross-compilation.

The gateway is ideal for:

• Multi-Model Apps: Simplify code by calling multiple models (GPT-4, Claude, Llama) via a unified interface.
• Cost Control: Automatically select the cheapest provider to reduce API costs.
• High Availability: Auto failover ensures service continuity for critical businesses.
• Rapid Prototyping: Use free/low-cost providers for development and switch to production-grade services without code changes.

Rust LLM API Router simplifies multi-LLM service management by unifying providers into one gateway. It combines Rust's performance with smart routing to balance cost, performance, and reliability. The project is open-source under MIT license, with detailed docs (CLI, API, architecture, routing) to support community contributions. It helps developers focus on business logic instead of infrastructure maintenance.