Show HN: Smart model routing directly in Claude, Codex and Cursor
40% cost savings. Weave just shipped a model router that intelligently picks the right AI model for each coding task—and it's available now.
Why it matters
As frontier model costs climb (Opus 4.7's tokenizer changes drove Weave's expenses up), intelligent model routing is becoming table-stakes for AI-first teams. This is a working solution shipping with real cost/quality data.
The key facts
6 to know40% token cost reduction with no quality loss
Routes requests across Opus 4.8, GPT 5.5, DeepSeek V4, GLM 5.2, Kimi K2.6
Trained RL routing model on tens of thousands of agent traces
Integrates with Claude Code, Codex, Cursor as drop-in endpoint
Source-available under Elastic License 2.0; hosted version at weaverouter.com
Built after Opus 4.7 tokenizer changes spiked internal AI coding costs
Go to the source
Hacker Newsgithub.com
Publisher excerpt: We built a model router that plugs into coding agents (e.g. Claude Code, Codex, Cursor, etc.) and intelligently sends requests to the best model to serve them. Here's a quick demo of running it locally: At Weave, we write ~all our code with AI, and it's been getting more expensive. This came to a…