The post explains that Jev is not an LLM but a decision model: give it options and it picks one, ask yes/no and it returns odds in milliseconds. While your main model writes, Jev decides what happens next, prompting discussion on model routing vs blocking risky agent commands.
Jev isn't an LLM. It's a decision model: give it options and it picks one, ask yes or no and it returns the odds, in milliseconds. Your main model writes, Jev decides what happens next. Where would you use it first, model routing or blocking risky agent commands?
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.
