The author describes Jev as handling filtering and decisions, letting the LLM go deeper only when worthwhile—an architecture for AI agents and workflows that prioritizes compute and cost.
Para visualizar mejor lo que contaba antes: Jev filtra y decide. El LLM profundiza solo cuando merece la pena. Esta separación me parece de las arquitecturas más interesantes para agentes y workflows de IA actualmente. Priorizando siempre compute y coste.
Gregor Zunic showcases an open-source browser agent combining Browser Use and Jev, finding flights in 7 seconds for $0.0039, using a new action space per step, DOM state space, and small LLM fallback for typing.
Scott Williams predicts custom Jev-style models (parallel constrained decoding) will make existing company agent systems more token efficient, creating millionaires.
The user praises the JEV model for its extremely fast reproduction of each movement and command, believing bots will become smarter as a result.
