Portable offline LLM knowledge system that runs in browser
Single-file RAG bundle runs entirely in browser without server setup.

Privacy-first hardware but unproven execution and unclear model quality vs. cloud alternatives.
Privacy-conscious professionals, creators, and organizations needing offline AI without cloud dependencies.
Jan.ai (open-source offline AI on local hardware) · Ollama (local LLM runner, free) · Apple Intelligence (on-device processing)
Single-file RAG bundle runs entirely in browser without server setup.
Two-stage local pipeline fixes filler words and grammar better than macOS dictation.
Four swap-able engines, 80ms latency, no subscription—beats Apple's 60-second timeout hard.
TikTok-style arXiv feed with on-device AI summaries, privacy-first, iOS-only.
Comprehensive offline AI suite (text, vision, images) on $300 phones—genuinely complete.
On-device Apple Intelligence summaries beat sending RFCs to the cloud.