Turn video and audio into text, offline.

One Rust binary extracts the audio track with ffmpeg and transcribes it with whisper.cpp — mp4, mkv, mov, wav, mp3 and more, with no cloud and no account.

Get started See it in action

4: output formats
~1.6 MB: single binary
0: cloud calls

SRT subtitles

$ transcribers talk.mp4 -f srt -o talk
wrote talk.srt

1
00:00:00,000 --> 00:00:05,200
Delete filler badges, drop redundant labels…

Everything happens on your machine.

Your media never leaves your disk. The model runs locally through whisper.cpp.

Fully local

No uploads, no API keys, no accounts. Audio is extracted and transcribed on your own hardware.

Four output formats

Plain text, SRT and WebVTT subtitles, or JSON with millisecond timestamps.

Models on demand

Pick tiny through large-v3 by name; missing models download once into a local cache.

One small binary

whisper.cpp is compiled in — about 1.6 MB, statically linked, only ffmpeg required.

ffmpeg pipeline

Any container ffmpeg reads — mp4, mkv, mov, wav, mp3 — is decoded to 16 kHz mono WAV before transcription.

Subtitle-ready

SRT and WebVTT output with millisecond timestamps, ready for players and video editors.

See it in action

Real CLI output — SRT subtitles straight to stdout or a file.

Transcribe to SRT

$ transcribers talk.mp4 -f srt -o talk
wrote talk.srt

1
00:00:00,000 --> 00:00:05,200
Delete filler badges, drop redundant labels, un-etalize the
random word in your header.

2
00:00:05,200 --> 00:00:12,640
Only use mono where necessary. Use a design system, show the
actual product, and large the main feature.

SRT to stdout (default)

$ transcribers talk.mp4
1
00:00:00,000 --> 00:00:05,200
Delete filler badges, drop redundant labels…

Pick a model

$ transcribers talk.mp4 --model small
downloading model ggml-small.bin ...
transcribing... 100%

Install in one command

Prebuilt binaries for macOS and Linux, or build from source with Cargo. Requires ffmpeg on PATH.

Install

$ curl -fsSL https://raw.githubusercontent.com/hieudoanm/transcribers/main/scripts/install.sh | bash
Installed to /usr/local/bin/transcribers
$ transcribers talk.mp4 -f srt -o talk
wrote talk.srt

Or from source: cargo install --git https://github.com/hieudoanm/transcribers transcribers (needs cmake for the whisper.cpp build). See the README for details.

View on GitHub