AI
Open-source Jev clones proliferate weeks after TypeSafe's launch
Ollaya launches as full-featured local decision-model framework
4
Sep 25 2:33 PM
· 2d ago · 1 post · 3 comments · 1 source · development 4 of 9
The most polished open-source implementation arrives with four model variants (laya, decider, von, qwen3guard), desktop and CLI tools, Docker support, and API compatibility with TypeSafe's SDKs. Achieves 10ms latency on RTX 4090 for five-question requests.
github.com
“Ask typed questions about any text or JSON and get calibrated answers in milliseconds. Private, open source, on your own hardware.”
Ollaya documentation TypeSafe Creator of Jev decision model Ollaya maintainers Open-source implementation leads TinyJev maintainers Open-source implementation leads Valen authors Multimodal extension developers
The whole story the bright band is this development · numbered dots are the others · click one to jump
Peak 8 pieces in two hours at Sep 21, 2 AM; 87 pieces over 7 days (3 articles · 17 posts · 67 comments)
Sep 21, 2 AM — 8 pieces · 2 articles · 2 posts · 4 comments — Hacker News 5, Newswires 2, Mastodon 1 Sep 21, 4 AM — 2 pieces · 1 post · 1 comment — Hacker News 1, Mastodon 1 Sep 21, 6 AM — quiet Sep 21, 8 AM — 5 pieces · 5 comments — Hacker News 5 Sep 21, 10 AM — 7 pieces · 1 post · 6 comments — Hacker News 6, Mastodon 1 Sep 21, 12 PM — 3 pieces · 3 comments — Hacker News 3 Sep 21, 2 PM — quiet Sep 21, 4 PM — 1 piece · 1 comment — Hacker News 1 Sep 21, 6 PM — 2 pieces · 2 comments — Hacker News 2 Sep 21, 8 PM — 2 pieces · 2 comments — Hacker News 2 Sep 21, 10 PM — 2 pieces · 2 comments — Hacker News 2 Sep 22, 12 AM — quiet Sep 22, 2 AM — 2 pieces · 2 comments — Hacker News 2 Sep 22, 4 AM — quiet Sep 22, 6 AM — quiet Sep 22, 8 AM — 1 piece · 1 post — Hacker News 1 Sep 22, 10 AM — 1 piece · 1 post — Hacker News 1 Sep 22, 12 PM — 1 piece · 1 post — Hacker News 1 Sep 22, 2 PM — quiet Sep 22, 4 PM — 1 piece · 1 post — Hacker News 1 Sep 22, 6 PM — quiet Sep 22, 8 PM — quiet Sep 22, 10 PM — quiet Sep 23, 12 AM — quiet Sep 23, 2 AM — quiet Sep 23, 4 AM — quiet Sep 23, 6 AM — quiet Sep 23, 8 AM — quiet Sep 23, 10 AM — quiet Sep 23, 12 PM — quiet Sep 23, 2 PM — quiet Sep 23, 4 PM — quiet Sep 23, 6 PM — quiet Sep 23, 8 PM — quiet Sep 23, 10 PM — quiet Sep 24, 12 AM — quiet Sep 24, 2 AM — quiet Sep 24, 4 AM — quiet Sep 24, 6 AM — quiet Sep 24, 8 AM — quiet Sep 24, 10 AM — quiet Sep 24, 12 PM — quiet Sep 24, 2 PM — quiet Sep 24, 4 PM — quiet Sep 24, 6 PM — 1 piece · 1 post — Hacker News 1 Sep 24, 8 PM — quiet Sep 24, 10 PM — quiet Sep 25, 12 AM — quiet Sep 25, 2 AM — quiet Sep 25, 4 AM — quiet Sep 25, 6 AM — quiet Sep 25, 8 AM — quiet Sep 25, 10 AM — quiet Sep 25, 12 PM — 1 piece · 1 post — Hacker News 1 Sep 25, 2 PM — 5 pieces · 5 comments — Hacker News 5 Sep 25, 4 PM — 8 pieces · 8 comments — Hacker News 8 Sep 25, 6 PM — 4 pieces · 4 comments — Hacker News 4 Sep 25, 8 PM — quiet Sep 25, 10 PM — quiet Yesterday, 12 AM — 2 pieces · 2 comments — Hacker News 2 Yesterday, 2 AM — 1 piece · 1 comment — Hacker News 1 Yesterday, 4 AM — quiet Yesterday, 6 AM — 2 pieces · 2 comments — Hacker News 2 Yesterday, 8 AM — quiet Yesterday, 10 AM — 5 pieces · 1 article · 1 post · 3 comments — Hacker News 4, Newswires 1 Yesterday, 12 PM — 1 piece · 1 comment — Hacker News 1 Yesterday, 2 PM — quiet Yesterday, 4 PM — 1 piece · 1 post — Hacker News 1 Yesterday, 6 PM — 1 piece · 1 post — Hacker News 1 Yesterday, 8 PM — 6 pieces · 1 post · 5 comments — Hacker News 5, Lobsters 1 Yesterday, 10 PM — 3 pieces · 3 comments — Hacker News 3 Today, 12 AM — 1 piece · 1 comment — Hacker News 1 Today, 2 AM — 1 piece · 1 comment — Hacker News 1 Today, 4 AM — 1 piece · 1 post — Hacker News 1 Today, 6 AM — 2 pieces · 1 post · 1 comment — Mastodon 1, Hacker News 1 Today, 8 AM — 1 piece · 1 comment — Hacker News 1 Today, 10 AM — quiet Today, 12 PM — 1 piece · 1 post — Hacker News 1 Today, 2 PM — quiet Today, 4 PM — 1 piece · 1 comment — Hacker News 1 Today, 6 PM — quiet 1–2 3 4–7 8 9
Sep 22 Sep 23 Sep 24 Sep 25 yesterday now · 7:37 PM ET
What people said 3 voices · verbatim
G
Has anyone actually seen better or the same results with Laya compared to Jev? From my experience, Laya performs significantly worse. It's less confident and often makes wrong decisions with more complex queries.
E
Cool... but this does seem undermined by the fact that Ollama can add support for decision models at any time.
R
>Run decision models locally.>example is a text classification task instead of a decision
All 9 developments of Open-source Jev clones proliferate weeks after TypeSafe's… →