4 posts

Tagged Local AI

All articles on the Glassy blog tagged Local AI.

WebGPU in 2026: running real AI models in your browser

Jun 27, 2026 · 7 min read

WebGPU plus WebAssembly has made in-browser AI a production reality. Real LLMs, real transcription, real embeddings — all running on your GPU, no install required. Here is the state of the art.

Your notes are not training data: privacy-first AI in practice

Jun 21, 2026 · 6 min read

Most AI note apps ship your content to a cloud vector database and ask you to trust their privacy policy. Glassy does the opposite: structural privacy, where there is nowhere for your data to go.

Local AI vs Cloud AI: what stays on your machine

Jun 13, 2026 · 5 min read

Glassy runs Qwen 2.5 and Whisper inside your browser via WebGPU. No API keys, no credits, nothing uploaded. Cloud AI is opt-in for the harder work. Here is the exact boundary.

Voice notes to searchable text: the Whisper pipeline

May 22, 2026 · 6 min read

Glassy Voice Studio records audio and transcribes it locally with Whisper running in your browser via WebGPU. No cloud API, no per-minute costs, no privacy concerns. Here is how the pipeline works.