Modus uses Jev for tool routing, precision rises from 44% to 83%
Modus hands each agent turn's tool and prompt-template choice to Jev, which returns calibrated probabilities in about 0.38 seconds and nearly doubles routing precision.

Jev from @TypeSafeAI is now running inside Modus. We've been testing it since launch, and the results blew past what we expected, so it's rolling out across the platform, on @RobinhoodCrypto, @solana, and every EVM chain we support. First place it landed: tool routing. Every agent turn decides which tools to load and which prompt template to use, and an LLM used to make that call, slowly and often wrong. Jev now makes it in ~0.38s with calibrated probabilities, and routing precision nearly doubled, from 44% to 83%. Benchmarks below👇