The latest release of diegosouzapw/omniroute is v3.8.51. The repository is a TypeScript-based, MIT-licensed gateway that connects multiple AI model providers behind a single endpoint. The provided FACTS contain no changelog or release notes for this version, so it is not possible to confirm what was added or fixed.
The repository description says it supports 359 providers and more than 1,200 models, including Kimi, Claude, GPT, Gemini, GLM, DeepSeek and MiniMax, and presents more than 150 of those providers as free. It highlights integrations with Claude Code, Codex, Cursor, OpenCode, Cline and Copilot, and states that it also supports automatic fallback routes to other providers in light of usage limits, along with MCP/A2A integration. Desktop and PWA deployment are also offered as features. The README claims that RTK and Caveman compression cut token usage by 15~95%.
The provider count and the token-saving rate are the repository's own claims, and the FACTS do not say whether they were independently verified. The free-tier catalogue published in the README contains 485 entries and 35 duplicate pools, and the stated estimate of 1.62 billion tokens per month is said to deduct those duplicate pools. That estimate is re-audited every two weeks and can move in either direction. It is therefore usable as a rough guide to features and scale, but the numbers alone make it hard to judge real cost savings or performance advantages.