chismesillo — local audio transcription

Drag & drop, or click to browseMP3, WAV, M4A, AAC, OGG, OPUS, FLAC, WebM — and video files too

Settings

Model

Bigger models are more accurate and slower to download and run.

Audio language

Task

Word-level timestamps

Slower, but gives a timestamp for every word.

Advanced

Chunk size (minutes)5

Audio is decoded and transcribed in chunks so memory stays flat on long files.

Force CPU (disable WebGPU)

Use this if your GPU produces errors or garbled text.

Preparing…

Segments will show up here as they are transcribed.

Export