Weave Router 2.0

Weave Router is an intelligent model router that dynamically directs each turn of a coding agent to the most cost-effective LLM capable of handling the task. It evaluates the complexity of every prompt using a trained classifier, then routes to a frontier model only when necessary, saving cheaper models for simpler tasks. The product is designed for engineering teams using AI coding assistants and agents, aiming to reduce inference costs by 30-60% while maintaining output quality. Weave Router integrates directly into existing agent workflows and provides detailed analytics on AI usage, code turnover, and quality metrics. In benchmarks, it achieved a high task completion rate at a significantly lower cost and faster per trial compared to a leading model. The platform offers a free Starter plan, a Pro plan at a set price per engineer per month, and custom Enterprise pricing. The company's broader mission is to make engineering work visible and measurable, helping teams optimize how they pair AI with human developers. Every customer gets a dedicated Slack channel, and pricing is transparent with no annual lock-in.