Thread

Kev: Tiny Jev-like family of decision models built on top of Qwen3.5

462 points · 211 comments · tosh

monkeydust · 1789977093 · focus · HN ↗
Bit of a Jev explosion going on. Is it because it's taking us back to a simpler time we understand better? Classification models have been around for a while.
1. Tycho · 1789979233 · focus · HN ↗
  It’s because it’s practically useful and enabled things that were impractical previously.
  1. petesergeant · 1789981148 · focus · HN ↗
    > and enabled things that were impractical previously
    I think that there are not _that_ many use-cases that have been opened up by this that tool-calling on other models didn't solve already. Really depends what benchmark you're looking at. This one against BANKING77[0] has many issues, but suggests it's really not far off DeepSeek 4.1 Flash. This one against BoolQ[1] shows marginal improvement over Qwen3.6. This one against MMLU-Pro[2] (same author as the previous) shows significant improvements over two Qwen models.
    So there's definitely _some_ alpha there, but I don't think it's the sea-change that the hype would suggest; that is to say, yes, some things that weren't practical before are now, but many things were already very practical with the existing tools.
    0: <a href="https://sanand0.github.io/llmevals/jev/" rel="nofollow">https://sanand0.github.io/llmevals/jev/
    1: <a href="https://github.com/ekzhang/openjev-sglang/blob/a3554ed9e9c26d5d7b3b2184a524fc61779dbcc1/evals/results/boolq-2026-09-18/comparison/report.md" rel="nofollow">https://github.com/ekzhang/openjev-sglang/blob/a3554ed9e9c26...
    2: <a href="https://github.com/ekzhang/openjev-sglang/blob/a3554ed9e9c26d5d7b3b2184a524fc61779dbcc1/evals/results/qwen38-27b/report.md" rel="nofollow">https://github.com/ekzhang/openjev-sglang/blob/a3554ed9e9c26...
    Fabricio20 · 1789995072 · focus · HN ↗
    The part about "tool-calling on other models didn't solve already" is what gets you, sure I could tool call deepseek, glm or any other model, but the latency is huge and you get no confidence score. I gave JEV a shot via OpenRouter and it has a reply in less than 400ms, it's fast enough and cheap enough that you can hook it up to a game loop for example (so highly state dependant) and it can do decisions in real time.
    petesergeant · 1790004583 · focus · HN ↗
    Please refer to the latency and price figures for DeepSeek V4.1 Flash in the first link I shared.

Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.