jevemu
MIT-licensed emulator of the Jev decision API over local vLLM models and hosted models that return logprobs, with a paired benchmark.
0 members list jevemu, unchanged over the last 12 weeks, data as of 16 November 2026
Members who list jevemu in their Stack, by week.
Members who list jevemu
Be the first to add jevemu
Nobody on Stackness runs jevemu yet. Add it to your stack and you're the first member running it.
About jevemu
Jevemu is an emulator that replicates the Jev decision API interface using local vLLM models and hosted models capable of returning log probabilities. Developers use it to test and benchmark applications built against the Jev API without depending on the hosted service, working instead with open-source language models they can run locally or on managed infrastructure. The tool includes a paired benchmark suite for comparing the performance of local and hosted models against the original Jev implementation.
Jevemu is developed by sinkarusa and distributed under the MIT license. The project is written in Python and available as open-source software, making it accessible for developers who want to integrate Jev-compatible functionality into their workflows or evaluate different model backends for decision-making tasks.
Checked 11 November 2026.
Nobody on Stackness runs jevemu yet.
Sign up and add itPosts about jevemu
- Do you need Jev? Typed decisions from an ordinary LLM with one token and logprobs
You can get Jev-style typed decisions from an ordinary LLM by forcing one output token and reading its logprobs. Where that matches Jev, and where it loses.