transformers-js

Browser LLM router: sends each prompt to the cheapest capable model tier and keeps a live savings ledger.

SAGI: fully local in-browser voice assistant (LFM2.5 LLM + Supertonic TTS on WebGPU via Transformers.js)

Read the water before you paddle out. Real swell, wind and verified tide turned into one explainable, sealed verdict per surf break.

Find any open tab by describing what it's about — private, on-device semantic tab search for Chrome (MV3)

Image search over CLIP embeddings with Lakebase Search

基于 Nuxt 4 的 AI 技术演示合集：MediaPipe / Transformers.js / WebLLM / Demucs，浏览器本地运行，含异步音频分离与 Python 参考实现

Mutes spoken profanity on YouTube, transcribed on-device with Whisper. No uploads, works offline. Chrome MV3 extension.

Clinical intelligence platform: RAG health assistant, medical report NLP, disease-risk models and a medication interaction checker, with patient, doctor and admin dashboards. Runs in your browser — try it on Hugging Face.

Give Claude the ability to watch local videos — extracts frames and transcribes audio entirely inside Node, with no external binaries

Fullstack AI-powered job matching platform. Scrapes Polish job portals, ranks offers using vector embeddings and cross-encoder re-ranking. Turborepo monorepo.

Transcribe video y audio a texto en tu navegador, sin subir el archivo a ningun servidor. Whisper local con transformers.js.

⚡ Serverless image+audio+text inference with Gemma 2, running 100% on-device via Chrome WebGPU. Zero backend GPU, $0.00/month.

Search your local images by text using CLIP embeddings (transformers.js) — shared image/text space, no API, no cloud.

Real-time Bangla & English meeting minutes. Whisper runs in the browser so audio never leaves your device; an incremental summariser builds decisions and action items live, each quoting its source line. FastAPI + Next.js + pgvector.

8 production-grade AI/ML projects: fine-tuned LLMs (DistilBERT, LLaMA, Code Llama, RoBERTa), RAG systems, NLP pipelines — each with FastAPI backend and live inference endpoints.

LucidLLM is a browser-based LLM chat application supporting local inference via WebGPU/WASM. It features multilingual support (i18n), session management, OPFS model storage, Google Drive backup, and real-time token speed monitoring. Built with vanilla JavaScript and Tailwind CSS.

Extensao Chrome (MV3) que transcreve audios do WhatsApp localmente via Whisper. O audio nunca sai da sua maquina.

An AI tab grouper for Chrome and Firefox that runs fully on your own computer

Open-source implementation of the typed-decision pattern popularised by TypeSafe's Jev: read a decision out of a small language model's logits, in the browser. Library, benchmarks and paper. Not affiliated with TypeSafe.