About Record to Notes
Keep full video transcripts in your own Notion, or transcribe local audio for free.
One complete task
Choose an audio file, get its transcript and take the text with you. Search, editing, timestamp playback and standard exports help you check the result before using it elsewhere.
Speech recognition runs on your device. This keeps server-side inference out of the free workflow, while device capacity still limits speed and supported recordings.
Open models
Speech recognition uses OpenAI Whisper (MIT). Speaker identification uses pyannote segmentation-3.0 (MIT) and the WeSpeaker ResNet34-LM speaker embedding model (CC BY 4.0), run in your browser through ONNX exports by onnx-community.
What is available today
The workbench accepts local audio files. Chapter navigation and extracted key points are optional aids to reading the transcript. Notion saving requires sign-in and is free when available. Cloud transcription and top-ups are not open yet.