Extremely fast and accurate audio transcriber surpassing Whisper. Fast on GPU or CPU.
whispr is a transcription pipeline built around OpenAI's Whisper model. It supports four backends (three interchangeable general-purpose ones plus an optional Mega-ASR specialist for degraded audio), parallel chunk processing, automatic fallback, and multiple output formats.
Open-source tool to transcribe any video — from lectures to tutorials — fully local and privacy-friendly.
Windows desktop app (Python, Tkinter) that transcribes live microphone audio to text with OpenAI Whisper running locally. No API key or cloud service; NVIDIA GPU optional. 🍂
Privacy-focused multi-platform audio recorder and AI meeting transcriber with speaker diarization — locally via whisper.cpp or cloud via ElevenLabs Scribe.
To associate your repository with the audio-transcriber topic, visit your repo's landing page and select "manage topics."