OpenMeetRec

OpenMeetRec records the audio of a video call, tab and microphone, then sends it to a speech-to-text API you configure to produce a Markdown transcript. It captures the tab's audio directly, so it works with any video conferencing site running in Chrome, not just one platform. No bot joins the call, and nothing is recorded until you click the extension icon. How it works: - Click the icon to start recording. A red dot and a REC badge show it's running, even after you close the popup. - Optionally get a reminder (a system notification) when you open a page that matches a video conferencing pattern you configured. Nothing starts automatically. - Audio is mixed from your microphone and the tab, then split into chunks while recording, so memory use does not grow with meeting length. - Each chunk is sent to your chosen provider (Mistral Voxtral or OpenAI Whisper, or a custom endpoint) as the meeting goes. A failed chunk can be retried without recording again. - The transcript merges cleanly across chunk boundaries and can group consecutive segments from the same speaker where the provider supports diarization. - The result is exported as a Markdown file. No server sits between your browser and the API you chose. The extension is scoped to transcription only. It does not summarize the meeting for you; that step is left to whatever tool you already use downstream. Permissions: capture uses activeTab, granted only for the tab you clicked the icon on. The optional meeting reminder needs the tabs permission to read that tab's URL (compared locally against your patterns, then discarded) and the notifications permission to show the reminder. Full breakdown: https://github.com/vlebert/openmeetrec/blob/master/docs/permissions-audit.md Open source, MIT license: https://github.com/vlebert/openmeetrec Privacy policy: https://github.com/vlebert/openmeetrec/blob/master/PRIVACY.md