Transcribe Audio

Local audio transcription using whisper-cpp — a C++ port of OpenAI's Whisper.

Quick Start

# Transcribe with distil-large-v3 (fast, high quality)
nix run nixpkgs#whisper-cpp -- -m models/ggml-distil-large-v3.bin -f audio.mp3

Model

Model	Size	Speed	Quality
ggml-distil-large-v3.bin	1.5 GB	⚡ Fast	Best

Download the Model

curl -L https://huggingface.co/distil-whisper/distil-large-v3-ggml/resolve/main/ggml-distil-large-v3.bin -o ggml-distil-large-v3.bin

Common Options

Option	Description
-m MODEL	Model path
-f FILE	Input audio file
-t N	Threads (default: 4)
-l LANG	Language (en, auto, etc.)
-otxt	Output to .txt file
-osrt	Output to .srt subtitle file
-ovtt	Output to .vtt file
-oj	Output to JSON
-of PATH	Output file path (without extension)
-nt	No timestamps in output
-np	No prints (results only)
--print-confidence	Show confidence scores

Examples

Transcribe with Timestamps (Default)

nix run nixpkgs#whisper-cpp -- -m ggml-distil-large-v3.bin -f recording.mp3

Save to Text File

nix run nixpkgs#whisper-cpp -- -m ggml-distil-large-v3.bin -f recording.mp3 -otxt -of transcript

Generate Subtitles

# First extract audio from video
ffmpeg -i video.mp4 -vn -acodec libmp3lame -ab 128k video.mp3
# Then transcribe
nix run nixpkgs#whisper-cpp -- -m ggml-distil-large-v3.bin -f video.mp3 -osrt -of captions

JSON Output with Confidence

nix run nixpkgs#whisper-cpp -- -m ggml-distil-large-v3.bin -f audio.wav -oj -of result --print-confidence

Auto-Detect Language

nix run nixpkgs#whisper-cpp -- -m ggml-distil-large-v3.bin -f audio.mp3 -l auto

Process Multiple Files

nix run nixpkgs#whisper-cpp -- -m ggml-distil-large-v3.bin -f file1.mp3 file2.mp3 file3.mp3

Offset and Duration

# Start at 30s, process 60 seconds
nix run nixpkgs#whisper-cpp -- -m ggml-distil-large-v3.bin -f audio.mp3 -ot 30000 -d 60000

Supported Formats

Audio: flac, mp3, ogg, wav

Video: Extract audio first:

ffmpeg -i video.mp4 -vn -acodec libmp3lame -ab 128k audio.mp3

GPU Acceleration

For GPU support, use whisper-cpp-vulkan:

nix run nixpkgs#whisper-cpp-vulkan -- -m ggml-distil-large-v3.bin -f audio.mp3