Reading up on Needle
2 deep · digging since sep 22
- Whistle: Speech to Text in 16.9 MB
Whistle is an open 16.9 MB speech‑recognition model that runs on CPU, transcribes seven languages in 11 ms first‑token latency, and integrates with the Needle engine for on‑device tool calls.
- awesome-llm-apps/advanced_llm_apps/needle at main · Shubhamsaboo/awesome-llm-apps
Needle is a local semantic search tool from the awesome-llm-apps repo that finds meaning-based matches in webpages or pasted text via a Chrome extension or React app, scoring sentences with TypeSafe Jev through Vercel AI Gateway and highlighting the most relevant original sentence rather than generating answers.