Open it in your browser and press Record. Text appears as people speak, each line is marked with who said it, and when you stop you get a written summary with the decisions and action items.
Free account. Works in Chrome, Edge, Firefox and Safari, on a laptop or a phone. Nothing to install.
DanaMorning. Can we get the pump order out this week?
MarcusYes, I'll send it to the supplier by Thursday.
PriyaAnd I'll check we have enough spare gaskets.
ScribANote gives you the words, who said each one, and what was agreed, so you can keep your attention on the conversation.
Text appears a second or two after each sentence. It splits the audio at natural pauses, so words aren't cut in half.
Each voice gets its own label. Name a speaker once, during or after the meeting, and every line from that voice updates.
When you stop, it writes a summary of the meeting with the decisions made and who owes what. You can tick off action items across all your meetings.
Every recording is saved. Click any line to hear that moment again, fix a misheard word, or regenerate the summary.
Your browser does the listening and our own servers do the rest.
Press Record and allow the microphone. Your browser listens for speech and sends a clip each time someone pauses.
A speech-recognition model (Whisper large-v3 turbo) on our GPU turns each clip into text, roughly twenty times faster than real time.
Each clip gets a voiceprint. The whole recording is regrouped as it goes, so early guesses about who is speaking get corrected.
When you stop, a language model reads the transcript and writes the summary, the decisions and the action items.
If you know who is in the room, set "People talking" to Exactly or Up to that number. It's the most reliable way to get speakers right, especially in the first minute.
The summary only uses a name if it was said aloud or you typed it in. Otherwise it writes "Speaker 2". It never makes a name up. Check the spelling of names before you send the notes to anyone.
Transcription, speaker recognition and summaries all run on hardware we operate. No third-party AI service receives your audio. Only you can see your recordings once you are signed in.
Create an account with your email address and start recording in under a minute.