Releases·Model Release
LibertAI releases Deem, an open-weight Jev alternative built on Qwen3.5-9B
LibertAI released Deem, an open-weight model positioned as an alternative to Jev, built on Qwen3.5-9B with a smaller 0.8B variant. It trails Jev on hard benchmarks but arrives early in a new model category, and the small version runs on CPU for local tinkering.
98 engagement1 sourceSat, Sep 26, 2026, 03:36 PM
AI Overview
- Deem 9B scores 68.9% on the hard benchmark with extended reasoning vs Jev's 74.1%, per LibertAI labs.libertai.io
- LibertAI also ships a 0.8B Deem version that runs on CPU for local tinkering
Entities
Related
Releases·Model Release
Kitani releases OpenJudgement-4B-Preview, an open-weight Qwen-based alternative to Jev / TypeSafe
A developer released OpenJudgement-4B-Preview, an experimental Qwen-based model fine-tuned on custom datasets for classification, scoring, and true/false judgments. It scores answer options directly rather than generating responses token by token, with Python formatting results into JSON with probabilities. The early preview is unfinished and positioned as a not-yet-Jev-level alternative, with the author soliciting feedback on failure cases.
A developer released OpenJudgement-4B-Preview, an experimental Qwen-based model fine-tuned on custom datasets for classification, scoring, and true/false judgments. It scores answer options directly rather than generating responses token by token, with Python formatting results into JSON with probabilities. The early preview is unfinished and positioned as a not-yet-Jev-level alternative, with the author soliciting feedback on failure cases.
23 engagement1 sourceredditThu, Sep 24, 2026, 02:19 PM
Social·Community
Qwen quietly ships decision-model-preview, a Jev competitor, with docs but no announcement
Reddit users report that Alibaba's Qwen team has published a docs page for a model called decision-model-preview, described as a competitor to Jev. The page appeared without any announcement, and no open weights are available yet.
Reddit users report that Alibaba's Qwen team has published a docs page for a model called decision-model-preview, described as a competitor to Jev. The page appeared without any announcement, and no open weights are available yet.
145 engagement1 sourceredditSun, Sep 27, 2026, 11:43 PM
Products·launched
jevos: open-source text classifier answering yes/no questions with a single forward pass
jevos is an open-source alternative to Jev: you send a text and a yes/no question and get an answer in one forward pass without generating text. It runs on a laptop CPU, making cheap yes/no classification over documents feasible. Aimed at developers who want fast, local rule-checking instead of full LLM generation.
jevos is an open-source alternative to Jev: you send a text and a yes/no question and get an answer in one forward pass without generating text. It runs on a laptop CPU, making cheap yes/no classification over documents feasible. Aimed at developers who want fast, local rule-checking instead of full LLM generation.
9 engagement1 sourceredditSat, Sep 26, 2026, 02:44 PM
Social·Community
Evaluating Open-source Jev alternatives with DecideBench
A practitioner evaluated open-source SystemOne models against Jev using DecideBench on a production-like dataset. No open-source model beat Jev, but several came close, with Imajev-4B reaching 95% accuracy as the cheapest model above 90%. The findings help teams pick cost-effective Jev substitutes.
A practitioner evaluated open-source SystemOne models against Jev using DecideBench on a production-like dataset. No open-source model beat Jev, but several came close, with Imajev-4B reaching 95% accuracy as the cheapest model above 90%. The findings help teams pick cost-effective Jev substitutes.
30 engagement1 sourceredditWed, Sep 30, 2026, 03:07 PM
Releases·Model Release
Kev: tiny Jev-like decision models (0.8B/4B/9B) on Qwen3.5 you can train and run locally - the 9B fits a 32GB Mac
Kev is a family of small Jev-architecture decision models (0.8B, 4B, 9B) built on Qwen3.5 and released under Apache-2.0 with training code and frozen eval suites. The 9B model runs in bf16 on a 32GB Mac, and the models are API-compatible with TypeSafe's System One, so its Python SDK works against a local Kev server unchanged. The release targets LLM-as-judge eval pipelines, and the web playground demonstrates option-order bias that judge models commonly exhibit.
Kev is a family of small Jev-architecture decision models (0.8B, 4B, 9B) built on Qwen3.5 and released under Apache-2.0 with training code and frozen eval suites. The 9B model runs in bf16 on a 32GB Mac, and the models are API-compatible with TypeSafe's System One, so its Python SDK works against a local Kev server unchanged. The release targets LLM-as-judge eval pipelines, and the web playground demonstrates option-order bias that judge models commonly exhibit.
34 engagement1 sourceredditTue, Sep 22, 2026, 03:08 AM