record audio in your browser.

what this is

a voice recorder built on the browser's native MediaRecorder API. you click record, your browser asks permission to use the microphone, and the audio captures into the page's memory. when you stop, you get a playable preview and a download button. no server is involved at any point.

the default capture format is webm/opus — what your browser supports natively, very compact for speech (about a megabyte per ten minutes). if your downstream tool needs something else, the page can convert to mp3 or wav using the same ffmpeg.wasm chain that powers our other tools. the conversion runs locally too.

nothing uploads

most online voice recorders pipe your audio to a server while you record. some advertise this as a feature ("save in the cloud"); some don't mention it but do it anyway. for voice memos, that's usually fine. for interview recordings, dictated chart notes, off-the-record passages, attorney-client calls, source-protected reporting — it's not.

this tool runs the entire capture path in your browser. open your network tab while you record; the only request is the page itself loading. nothing leaves the tab. when you close the tab, the recording goes with it unless you've downloaded it. for the structural argument and how to verify this, see this post.

how to use it

1. click start recording and approve the microphone permission your browser asks for. (you can revoke it anytime from the site settings.)
2. speak. the page shows elapsed time and bytes captured so far. recording continues until you click stop or close the tab. there's no per-recording length cap from the browser side, but past about 60 minutes the in-memory file gets unwieldy on lower-end machines.
3. click stop. the recording finalizes and a playable preview appears. listen back, decide whether you're keeping it.
4. choose a format and download. webm is the no-conversion option (instant). mp3 and wav each run a local ffmpeg transcode, taking a few seconds per minute of audio.

which format to pick

• webm — smallest file, best for archiving voice memos. accepted by most modern transcription tools (including audiohighlight) without conversion.
• mp3 — universally compatible. if you're uploading to an older transcription service or to a podcast host, this is the safe choice.
• wav — uncompressed PCM. larger files, but lossless. archival recordings, broadcast ingest, audio-engineering follow-ons.

what it can't do

• multi-track recording. MediaRecorder captures one channel at a time. for a two-mic interview where each speaker should be on a separate track, use a dedicated DAW (Logic, Reaper, Audacity). that's a different category of tool.
• system-audio capture. this tool records from the microphone only, not from "what's playing on your computer." for that you need system-level capture (Loopback on mac, VoiceMeeter on windows) or screen-share with audio.
• save-as-you-go to disk. the recording lives in browser memory until you download it. if you're capturing a 90-minute interview on a low-memory machine, monitor the tab — at very long durations, browsers can OOM.

browser support

chrome, edge, firefox, safari, arc — all current desktop versions support MediaRecorder. the codec choice differs (chrome and firefox produce webm/opus; safari produces mp4/aac), but the page handles that. mobile browsers also work, though file management on mobile is fiddly.

what to do with the recording

depends on the job:

• transcribe it. audiohighlight's transcription workspace is the obvious follow-on. drop the recording in for a transcript with bulk speaker-fix and click-word-to-replay-audio.
• extract or trim a clip. run it through extract-audio-from-video or the audio-trimmer (shipping next) to crop down to the part that matters.
• publish as a podcast or voice memo. webm and mp3 both upload cleanly to most podcast hosts and social platforms. wav for archival.