scriberr

Scriberr is an open-source, completely offline audio transcription application designed for self-hosters who value privacy and performance. It allows local transcription of audio and video on your own machine, ensuring no data is sent to third-party cloud providers. ### Key Features - **Local & Private Transcription**: Transcribes audio and video locally using state-of-the-art machine learning models such as NVIDIA Parakeet, Canary, and Whisper, delivering high-accuracy text with word-level timing. - **Smart Speaker Detection**: Automatically detects different speakers (diarization) and labels exactly who said what. - **Chat with your Audio**: Integrates with Ollama or OpenAI API-compatible providers to generate summaries, ask questions, or hold conversations with transcripts inside the app. - **Workflow Integration**: Features an extensive API and a Folder Watcher that automatically processes new files in a folder, fitting easily into existing automations like n8n. - **Capture & Organize**: Includes a built-in audio recorder for on-the-go voice memos and integrated note-taking features to annotate transcripts while listening. - **PWA Support**: Supports Progressive Web App installation, providing a native app experience on both desktop and mobile devices. ### Installation & Configuration Scriberr offers multiple installation methods: - **Homebrew**: Easy installation on macOS and Linux via `brew install scriberr`. - **Docker**: Offers standard CPU Docker images, CUDA-optimized images for NVIDIA GPUs, and a Blackwell-specific image for RTX 50-series GPUs. Configuration can be customized using environment variables (e.g., `PORT`, `HOST`, `DATABASE_PATH`, `OPENAI_API_KEY`, `JWT_SECRET`). ### Project Status The project's active development is currently paused due to the creator's layoff, but it is not abandoned. The creator welcomes community contributions and collaboration to help move the project forward during this period.