InfoHacker News · community
Show HN: Open-source model routing for coding agents at Astra-level performance
Weave released Weave Router 2.0, an open-source model router for coding agents that intelligently switches between LLMs. On Terminal Bench 4.0 it matched GPT-6 Astra's pass rate at 52% of the cost and 2.2x faster completion; on SWE Atlas it matched pass rates at 54% of the cost and 2.5x faster. The improvements come from a new hidden Markov model plus classifier architecture, a larger training dataset, and smarter cache-eviction impact calculation.
Read the full story on news.ycombinator.com
What to doTry the open-source Weave Router if you want to reduce costs and latency for coding agents by routing between multiple LLMs instead of using a single frontier model.
- Source
- news.ycombinator.com