In simple English
This tool turns your recorded audio and video files into written text on your own computer. It keeps your files private so nothing leaves your server. It can tell different speakers apart and label who spoke each line. You can also chat with your transcripts using a local language model. It is completely free to run, though running local speech models needs a computer with a decent graphics card or plenty of memory.
What it is
A self-hosted audio transcription server written in Go with a web interface, using Whisper (a computer program that turns speech into text) and other speech models. It provides smart speaker detection and folder monitoring for automatic processing.
What it is good for
Solo builders who want to transcribe voice memos, interviews or meeting recordings without paying for a cloud subscription or sending private data to a third-party service.
- Runs entirely on your own hardware without sending data to the cloud.
- Detects multiple speakers and labels who said what in your recordings.
- Includes an application programming interface for connecting with workflow tools.
- Requires an NVIDIA graphics card for fast transcription on large files.
- First startup takes several minutes to download large speech and voice models.
- Running local language models for chat needs additional setup with tools like Ollama.
Paid cloud transcription services like Plaud or Otter
Ollama for running local chat models alongside transcripts
- openai/whisper111k stars
The underlying speech recognition model, but provided as a command-line tool without a web interface.
- ggml-org/whisper.cpp55k stars
A fast C++ port of Whisper for local transcription, lacking the built-in web dashboard and speaker detection of Scriberr.
- mpv-player/mpv38k stars
A command-line media player that handles most audio, video and subtitle formats and can be driven entirely from scripts.
- snapcast/snapcast8.0k stars
A server and clients that play one audio stream in several rooms, kept in time with each other.
- MusicPlayerDaemon/MPD2.8k stars
A background music player with no interface of its own, controlled over a network by separate client programs.
docker compose up -d