Stars
Forks
Watchers
ScribeWizard
Audio lectures become structured, Markdown-formatted notes in about a minute with ScribeWizard (also known as GroqNotes). Upload an MP3, WAV, or M4A file - or paste a YouTube link - and the app runs a three-stage pipeline on Groq's LPU inference hardware: Whisper Large v3 transcribes the audio, a larger Llama model drafts a comprehensive outline of the material, and a faster Llama model fills each section with detailed content. This scaffolded prompting strategy is the core idea: the strong model handles structure where quality matters most, the fast model handles volume, and Groq's 1200+ tokens-per-second inference keeps the whole process near real time. Output renders as clean Markdown with support for tables and code blocks, and finished notes download as text or PDF. Model selection is configurable - swap in other Groq-hosted open models like Mixtral or Gemma to trade speed against quality or work around rate limits. Built as a single Streamlit app by Benjamin Klieger at Groq, it needs only a Groq API key to run, making it one of the simplest self-hosted AI tools to operate.
Benefits
- An Hour of Audio, a Minute of Waiting
- Groq's LPU inference runs transcription and generation at 1200+ tokens per second, so full lectures become finished notes while conventional tools are still transcribing.
- Notes, Not Transcripts
- The outline-then-content pipeline produces organized study material with headings, tables, and code blocks - a different artifact than the raw text dump most transcription services return.
- Pay Per Use, Not Per Seat
- Transcription SaaS charges monthly per user. A self-hosted instance costs only Groq API usage, and one deployment serves the whole team.
- Audio Stays Under Your Control
- Meetings and lectures upload to your own instance rather than a note-taking vendor's cloud, with only inference calls leaving the server.
Features
- Audio and YouTube Input
- Upload MP3, WAV, or M4A files up to 100 MB, or paste a YouTube link for direct processing.
- Whisper Large v3 Transcription
- High-accuracy speech-to-text running on Groq for near-instant turnaround.
- Scaffolded Dual-Model Generation
- A stronger Llama model plans the outline; a faster one writes section content - balancing quality, speed, and rate limits.
- Markdown Output with Tables and Code
- Notes render with clean formatting including tables and syntax-highlighted code blocks.
- Text and PDF Export
- Download the complete notes as a text file or styled PDF.
- Configurable Models
- Swap the outline and content models for other Groq-hosted options like Mixtral or Gemma.