{"slug": "show-hn-snap-local-first-decision-engine-open-weight", "title": "Show HN: Snap – local first decision engine open weight", "summary": "Developer emnlmn released Snap, an open-weight, MIT-licensed local decision engine positioned as a drop-in replacement for the TypeSafe System One API, running the 2B-parameter snap1-2b model in a single forward pass to classify, rank and decide on inputs in sub-second time on a local machine. Snap is installable on macOS via `brew install emnlmn/snap/snap` followed by `snap serve`, which pulls snap1-2b from Hugging Face and starts a local web server and playground UI. The developer reports a 0.643 zero-shot accuracy score on zero-shot typed-decisions and asks the community for feedback on the runtime and on training snap1.", "body_md": "I've built a local decision engine drop-in replacement for the TypeSafe System One api because i wanted to test locally some sensitive use cases and at the time jev subscriptions was closed.\n\nSnap supports 4 class of questions that let you classify, rank and decide about an input in sub-second on your local machine. It doesn't generate text, it reads logits in a a single forward pass.\n\nIf you have a mac you can run brew install emnlmn/snap/snap and then snap serve. It will pull the snap1-2b model from HF and start a local web server and web playground ui.\n\nEven if the benchmarks show a 0.643 zero-shot accuracy score on zero-shot typed-decisions, snap1-2b it's 2b model, you can't expect high intelligence from it.\n\nI'd like to share it with the community and would love to get feedback about the runtime and the training i've done with snap1. It's my first big work in the machine learning world and i hope it can be useful to someone that want to test out decision model locally.\n\nOpen weight model, MIT license.\n\nComments URL: [https://news.ycombinator.com/item?id=49921892](https://news.ycombinator.com/item?id=49921892)\n\nPoints: 1\n\n# Comments: 0", "url": "https://wpnews.pro/news/show-hn-snap-local-first-decision-engine-open-weight", "canonical_source": "https://emnlmn.github.io/snap/", "published_at": "2026-10-01 14:06:18+00:00", "updated_at": "2026-10-01 14:18:33.408504+00:00", "lang": "en", "topics": ["artificial-intelligence", "machine-learning", "ai-tools", "ai-products", "large-language-models"], "entities": ["Snap", "snap1-2b", "emnlmn", "TypeSafe System One", "Hugging Face"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/show-hn-snap-local-first-decision-engine-open-weight", "markdown": "https://wpnews.pro/news/show-hn-snap-local-first-decision-engine-open-weight.md", "text": "https://wpnews.pro/news/show-hn-snap-local-first-decision-engine-open-weight.txt", "jsonld": "https://wpnews.pro/news/show-hn-snap-local-first-decision-engine-open-weight.jsonld"}}