transcriptionstream/transcriptionstream
A self-hosted offline service that transcribes audio files using Whisper and generates summaries via Ollama/Mistral LLM.
Not currently ranked — collecting fresh signals.
star history
Transcription Stream provides a complete pipeline for converting audio to text with speaker diarization. It uses Whisper and WhisperX models for speech recognition, Ollama with Mistral-7B for LLM-powered summarization, and Meilisearch for full-text search of transcripts. The service offers both a web interface and SSH drop zones for file upload, making it a turnkey solution for offline audio processing.
Frequently asked
- What is transcriptionstream/transcriptionstream?
- A self-hosted offline service that transcribes audio files using Whisper and generates summaries via Ollama/Mistral LLM.
- Is transcriptionstream open source?
- Yes — transcriptionstream/transcriptionstream is open source, released under the GPL-3.0 license.
- What language is transcriptionstream written in?
- transcriptionstream/transcriptionstream is primarily written in Python.
- How popular is transcriptionstream?
- transcriptionstream/transcriptionstream has 944 stars on GitHub.
- Where can I find transcriptionstream?
- transcriptionstream/transcriptionstream is on GitHub at https://github.com/transcriptionstream/transcriptionstream.