ej: an 11 MB open model for typed decisions in one pass
ej 0.0.1 is an open 11.4 MB System-One decision model (Apache-2.0 code, weights on Hugging Face) that answers typed choice/yes-no/score questions in one forward pass with calibrated probabilities — no token generation, CPU-only, 214 ms warm per record.
Why it matters
Tiny, auditable decision primitive with fitted calibration, few-shot workflow adaptation, and an unusually rigorous benchmark against Jev, laya, kev and others on the same suites.
What you could build with it
An indie developer automating support or ops workflows could replace fragile LLM prompt-and-parse routing with ej's deterministic 11 MB CPU call, using the per-question probabilities to automate only high-confidence decisions and route the rest to humans.
Does it hold up?
Too early to judge — v0.0.1 landed 2026-10-09; the author is candid that accuracy on unseen workflows is low (0.42 group-macro) and urges measuring on your own labeled cases before automating.
Learn more
First spotted on github: source.
More AI models
OpenAI publishes 722 AI-generated math manuscripts from an unreleased modelOpenAI released 722 mathematical manuscripts in 372 result families to GitHub under Apache-2.0, produced by an…models · JEV 0.74Mistral Large 4 (Le Chonk): 1.05T-param open-weight MoEMistral's new flagship: a 1.05T-parameter mixture-of-experts multimodal model with 49B active params and 1M-token…models · JEV 0.72ElevenLabs Eleven v4 + v4 Turbo: new emotive TTS architecture with a 100ms real-time variantElevenLabs released Eleven v4 and Eleven v4 Turbo on September 28: a new text-to-speech architecture with inline…models · JEV 0.72VoiceStudio: open-source, fully-local ElevenLabs alternative with voice cloning, dubbing and dictation in 646 languagesVoiceStudio is an open-source, fully-local voice AI studio: voice cloning from a 3-second sample, voice design, video…models · JEV 0.71JetBrains Mellum 2.1: 12B MoE coding model trained for agentsJetBrains released Mellum 2.1, a 12B MoE model (2.5B active parameters, 128K context) under Apache 2.0, trained mainly…models · JEV 0.69DeepSeek V4-Flash official API: public beta with upgraded agent capabilities, Responses API and Codex supportDeepSeek's official V4-Flash is now a public beta API with massively upgraded agent capabilities - benchmark scores…models · JEV 0.69
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.