Show HN: Teach a speech model (en) a name by saying it once, on your Mac

Category: ai-ml

Tags: speech-recognition, local-inference, macos, fine-tuning, privacy

Score: 6.7/10 (Innovation: 7, Technical: 7, Documentation: 5, Utility: 6)

Detail Keeper is a local speech-to-text tool for Mac that lets you teach a name by speaking it once, using a fine-tuned Qwen3-ASR adapter. It avoids cloud processing and claims competitive accuracy on name spelling versus ElevenLabs and Deepgram, with published benchmark results.

Target audience: Mac users, privacy-conscious professionals, and developers needing customizable local ASR

Repository: https://detailkeeper.com/

View on Hacker News