workweave/router

Model router for agentic systems. Routes every prompt to the right model in <50ms. Cut costs 40-70% with just an endpoint change.

What it solves

It eliminates the need to manually choose an LLM for every task by providing a single API endpoint that automatically routes requests to the most suitable model. This prevents the "vibes-based" approach to model selection, ensuring that the best model is picked for each specific action based on performance data rather than intuition.

How it works

The router acts as a drop-in proxy for major providers (Anthropic, OpenAI, Gemini) and OpenAI-compatible endpoints (like OpenRouter for DeepSeek, Llama, and Mistral). It uses a tiny on-box embedder and a cluster scorer derived from the Avengers-Pro research to evaluate each upstream API request and route it to the optimal model in real-time. It supports streaming, tools, and vision across these providers.

Who it’s for

It is designed for developers using AI-powered coding tools (such as Claude Code, Cursor, or Codex) or those building their own applications who want to optimize model performance and cost without changing their codebase for every model update.

Highlights

  • Multi-Provider Support: Compatible with Anthropic, OpenAI, Gemini, and various OSS models via OpenRouter.
  • Action-Based Routing: Routes per individual action rather than per conversation turn.
  • Drop-in Proxy: Can be integrated into existing tools by simply pointing them to a local or hosted endpoint.
  • Security: Provider keys are stored locally and encrypted at rest.
  • Observability: Built-in OTLP traces for integration with dashboards like Weave, Honeycomb, or Datadog.

Related

  • Project
    Mirrowel/LLM-API-Key-Proxy

    A self-hosted universal LLM API proxy that provides OpenAI and Anthropic compatible endpoints for multiple providers with built-in key rotation and failover.

  • Project
    Continuum-AI-Corp/OrcaRouter-Lite

    A self-hosted, OpenAI-compatible LLM router that automatically routes requests to the cheapest capable model across multiple providers.

  • Project
    ulab-uiuc/LLMRouter

    An open-source library for intelligent LLM routing that dynamically selects the optimal model for each query to balance cost and performance.

  • Project
    NadirRouter/NadirClaw

    A local LLM routing proxy that cuts API costs by 40-70% by routing prompts to the cheapest viable model and escalating to premium models only when necessary.

  • Dispatch
    Weave Router: Intelligent Model Routing for Agentic Systems

    Weave Router is a drop-in proxy that reduces LLM costs by 40-70% by routing prompts to the most efficient model in under 50ms using a cluster scorer derived from Avengers-Pro.