Scriberr

Scriberr is an open-source, self-hosted audio transcription application that runs completely offline on your own hardware. Using state-of-the-art ML models like NVIDIA Parakeet, Canary, and Whisper, it delivers high-accuracy transcripts with word-level timing. Key features include automatic speaker detection (diarization), AI chat with your transcripts via Ollama or OpenAI API, a built-in audio recorder, folder watcher for automated processing, and a polished PWA interface working on both desktop and mobile. Built in Go, Scriberr ensures no audio data ever reaches third-party cloud services. Perfect for journalists, researchers, students, and privacy-conscious users who need accurate transcription without subscription fees. Self-host it with a single Docker command and maintain full control over your data.