Radar / AI infrastructure / OpenRouter Auto Router
OpenRouter Auto Router: next-gen model routing with a dedicated coding router
OpenRouter's rebuilt Auto Router picks the model and reasoning effort per request, alongside a new Auto Code router tuned for coding agents.
Why it matters
Auto Code hits 74.3% on Deep SWE at $1.98 per task, roughly 74% cheaper than single frontier-model calls at comparable accuracy, with per-request routing rationale exposed.
What you could build with it
A startup running a coding agent could point its harness at openrouter/auto-code and let the router pick the cheapest capable model per task, cutting inference spend while keeping a per-request rationale for debugging bad picks.
Does it hold up?
Too early to judge: launched 6 Nov. OpenRouter reports 2.5T tokens a month through Auto Router, but the new generation's real-world cost and quality trade-offs are not yet independently measured.
Built with OpenRouter Auto Router
- Introducing the New Auto and Auto Code Routersopenrouter · OpenRouter's launch announcement: the new Auto Router weighs eleven factors per request and the coding-specialized Auto Code router is built for multi-turn software engineering and agentic tool use.
- OpenRouter Launches Next-Generation Auto and Auto Code Model Routersheadsupai · HeadsUpAI news tracker on the launch: Auto Code benchmarks similar to Claude Opus 5.5 at 74% lower cost per task, with a 0.65s p90 routing decision.
Learn more
First spotted on article: source.
More AI infrastructure
colibri: run 744B-parameter MoE models on consumer hardwarePure-C inference engine with zero dependencies that streams MoE experts from NVMe disk, keeping hot experts on GPU…infra · JEV 0.78claude-mem: persistent memory plugin that lets coding agents pick up where the last session left offPlugin that hooks five Claude Code lifecycle events, compresses session activity into a local SQLite + Chroma store…infra · JEV 0.74FreeToken: datacenter-scale MoE serving on desktop GPUsAn edge-native mixture-of-experts serving engine from UC Berkeley researchers that runs 290B+ parameter MoE models on…infra · JEV 0.74XLA Serving Engine: Google open-sources its C++ inference runtimeGoogle's general-purpose model export and inference runtime for the OpenXLA ecosystem, now open source under Apache 2.0.infra · JEV 0.67TorchTPU: Google's native PyTorch stack for TPUs, now open sourceAnnounced at PyTorch Conference 2026 on October 20: Google open-sourced TorchTPU, its native PyTorch backend for Cloud…infra · JEV 0.66billion-context: month-long agent sessions in a 100K context windowA local proxy that sits between a coding agent and its model API, letting the model fold old messages into searchable…infra · JEV 0.65
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.