Voice Note Transcriber

What it does

Record audio and transcribe it at the same time, with the live transcript appearing word-by-word as you speak. When you stop, you've got both: the original audio file (downloadable WebM) and the time-stamped transcript (copyable text). Useful when you want a record of both — what was said AND what your microphone actually heard, rather than picking one and trusting it.

Common uses: quick meeting notes (record the call, have a transcript to scan afterward instead of re-watching); voice memo capture (record yourself thinking out loud, get text you can search later); interview prep / debrief (transcribe a phone screen or fact-finding call); language practice (read aloud, see what the recognizer thinks you said — exposes mispronunciations and accent issues).

Recording uses MediaRecorder for the audio (output: WebM with Opus codec). Transcription uses the browser's built-in SpeechRecognition API which on Chrome / Edge / Brave routes audio through Google's hosted recognition service. So while the audio file stays in your browser, the recognition (transcript) DOES require an internet connection and your audio is briefly transmitted to Google for processing. If full local-only recording is required (e.g. confidential meetings), use a desktop tool with Whisper running locally.

How to use it

1. Pick your spoken language from the dropdown. Wrong language picks produce garbled transcripts.
2. Click Start. Browser asks for microphone permission the first time — grant it.
3. Speak naturally. Words appear word-by-word in the transcript pane; the audio waveform shows your input level.
4. Click Stop when you're done. The transcript box is editable — fix any recognition errors before copying.
5. Two outputs: Download Audio gives you a WebM file; Copy Transcript puts the text on your clipboard. Or download both as a paired set.

When to use this tool

• Recording quick voice notes where you want both audio and a searchable transcript.
• Capturing meeting audio + a rough transcript without a paid transcription service.
• Practicing speeches and reviewing both the recording and what the recognizer thinks you said.
• Language learning — read aloud, compare the transcript to your intended text.

When not to use it

• Confidential / sensitive content — Chrome's transcription routes audio through Google's recognition service. For local-only transcription, use a Whisper-based desktop tool.
• Multi-speaker transcripts — the API doesn't separate speakers. Use Otter, Rev, Whisper-based diarization tools for proper speaker turns.
• Long recordings (>60 minutes) — browser memory is a bottleneck.
• Production-grade transcription — accuracy is ~85-95% in good conditions; for legal, medical, or court-quality transcripts use a professional service.