diegosouzapw/OmniRoute

Never stop coding. Free MIT AI gateway: one endpoint, 290+ providers (90+ free), 500+ models — Kimi, Claude, GPT, OpenAI, Gemini, GLM, DeepSeek, MiniMax. Works with Claude Code, Codex, Cursor, OpenCode, Cline & Copilot. Quota-aware auto-fallback, RTK+Caveman compression saves 15-95% tokens, MCP/A2A, Desktop/PWA. Built by 500+ contributors

What it solves

OmniRoute is an AI gateway that prevents coding interruptions caused by API rate limits, expiring quotas, and high costs. It aggregates hundreds of AI providers and free tiers into a single endpoint, allowing users to switch between models and providers automatically without changing their tool configuration.

How it works

OmniRoute acts as a local proxy (running on localhost:20128) that translates requests from OpenAI-compatible tools (like Cursor, Claude Code, or Cline) to a vast network of 290 providers. It uses a "tier cascade" system to route requests through four levels: Subscription, API Key, Cheap, and Free. If a provider in a higher tier fails or hits a limit, it silently falls back to the next available option.

Who it’s for

It is designed for developers and AI power users who use multiple coding agents and LLMs and want to maximize their free tiers, avoid surprise bills, and maintain zero downtime during development.

Highlights

  • Massive Provider Network: Access to 290 providers, including over 90 free tiers and 516 models.
  • Automatic Fallback: 19 different routing strategies (e.g., cost-optimized, latency-first, or quality-first) to ensure requests always get a response.
  • Token Compression: Uses RTK and Caveman stacked compression to reduce token usage by an average of 89% (ranging from 15% to 95%).
  • Zero Config Start: Works immediately upon installation with pre-wired keyless free providers.
  • Local-First Privacy: Features AES-256-GCM encrypted keys and local-first architecture.