AnyJev
Turn any LLM into a Jev-style decision model: typed decisions, real probabilities, no training. (continue updating, welcome any issue and PR request)
AnyJev launched on 21 September 2026. It ranked #58 on Daily Launch on 17 October 2026 with a Launch Score of 31, from +67 stars on 17 Oct on GitHub. It was new in the ranking that day.
What is AnyJev?
kind is "choice" (default), "yesno" or "score" (options are ordered levels, lowest first, numbered from 1; firstlevel=0 numbers them from 0). decidebatch([...]) takes a list of such dicts. Send the hardest decisions to reasoning, with a cap. A decision whose top two options are less than tau apart in log-probability is answered again with thinking on. After the reasoning, the answer is read as a label distribution, so it also has probabilities and cannot fail to parse.
From its README on GitHub
AnyJev is a repo by nokia-applied-research on GitHub. It is written in Python and released under the Apache-2.0 license.
How to get AnyJev
Get it from its repository on GitHub.
Launch timeline
Launch Score, last 7 days
Ranked on one day so far: 31 of 100 on 17 October.