Show HN: Jeva.cpp – a llama.cpp fork with JEV-compatible API for all LLMs Jeva.cpp, a fork of llama.cpp, adds a JEV-compatible decision API to llama-server that enables Choice, Score and Noul evaluations directly from model logits while preserving standard autoregressive generation. The fork works with all models and platforms supported by llama.cpp, reusing its existing model implementations and inference backends, though the JEV decision API requires models that provide next-token vocabulary logits. jeva.cpp is a fork of llama.cpp that adds a JEV-compatible decision API https://github.com/PragmaTwice/jeva.cpp/blob/master/docs/jeva.md to llama-server , enabling Choice, Score and Noul evaluations directly from model logits while preserving standard autoregressive generation. jeva.cpp is designed to work with all models and platforms supported by llama.cpp, reusing its existing model implementations and inference backends. The JEV decision API requires models that provide next-token vocabulary logits; other model types retain their original functionality. LLM inference in C/C++ ggml https://github.com/ggml-org/ggml / ops https://github.com/ggml-org/llama.cpp/blob/master/docs/ops.md / maintainer PRs