local-whisper-transcriber

Turns speech in an audio or video file already saved on the user's machine into a timestamped transcript using ffmpeg and whisper.cpp, keeping conversion, inference and transcript storage on the local disk with no cloud speech service. Chooses a model by language, including tuned Chinese and Japanese variants, verifies or fetches GGML weights from Hugging Face with checksum checks, reuses earlier transcripts unless a fresh pass is requested, writes a JSON transcript with time-stamped segments plus a plain text copy, and passes summary work to a summarization skill. Fetching a missing model is the only step that needs internet access. Use when the user wants a transcript, captions or speech to text for a local file without cloud services. Not for downloading media from streaming platforms, live microphone capture, or writing summaries itself.

Version

v1.0.0

Size

17 KB

Updated

2026-10-10

Your client verifies the signature before reading. This explicit manual read is available on every plan. Free automatically loads five chosen skills; paid plans automatically load relevant skills across the library.