Till JS 631cdafdb5 feat(voice): route STT through local Whisper when model is loaded ... transcribeAudio() now checks localSTT.isReady before falling back to the server-side mana-stt proxy. When local STT is active, audio blobs are decoded to Float32Array via AudioContext.decodeAudioData() and transcribed entirely on-device. The returned model field shows "Whisper Tiny (lokal)" or similar so every module (dreams, memoro, habits) displays which backend was used — no module code changed. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>		2026-04-12 16:17:56 +02:00
..
landing	refactor(mana): rename inventar → inventory across the codebase	2026-04-09 15:50:24 +02:00
mobile	chore: complete ManaCore → Mana rename (docs, go modules, plists, images)	2026-04-07 12:26:10 +02:00
web	feat(voice): route STT through local Whisper when model is loaded	2026-04-12 16:17:56 +02:00