An online voice recorder lets you capture your microphone in a browser tab and save the audio to your own device. You press record, allow the microphone once, pause when you need to, then download the take. No installer, no account and no upload are involved.
The recorder is a thin layer over two browser interfaces: getUserMedia opens the microphone, and MediaRecorder encodes whatever it hears, entirely inside the tab. Nothing is sent to a server at any point, including during the optional conversion to MP3, which runs in a WebAssembly build of FFmpeg on your own machine. The price of that design is a format you do not fully control: the container is chosen by the browser, Chrome and Edge usually write WebM with Opus, Safari writes MP4 with AAC, and the download extension follows. The recorder also asks for nothing until you click, so opening the page does not touch your microphone.
A quick voice memo, a dictated draft, a spoken homework answer, an interview clip: all of these need a microphone and nothing else, and a browser tab is the shortest path from that idea to a finished file. You do not install anything, and you do not hand the recording to someone else's server just to get an audio file back. What you give up in exchange is control over the container, and this guide names that limit and the others before you hit record rather than after.
What a browser voice recorder actually captures
Everything in the finished file comes from one device, the microphone, behind one permission prompt. The page asks for audio only, so no camera is involved and there is no video track to strip out later. Until you allow that prompt the recorder has no input at all, and the level meter stays flat.
The recorder opens the microphone exactly as the browser hands it over and encodes one continuous stream. There is no second track, no backing music and no system audio, which is why this is the wrong tool for capturing a call that plays through your speakers.
| What lands in the file | Included | Detail |
|---|---|---|
| Your voice from the microphone | Yes | Audio only, one permission prompt |
| System or tab audio | No | The recorder has no access to it |
| A second track or overdub | No | One stream per take |
| Watermark or beep | No | The page adds nothing to the audio |
| Trimming or editing | No | One straight take, cut later |
| Upload to a server | No | Audio stays in the tab until you download |
Voice Recorder
Open the microphone, watch the level meter, record and pause, then download the take or convert it to MP3. No install and no watermark.
How to record audio online
Three actions, and only the first one needs a decision.
- Choose the input and allow access. Pick a microphone from the list if you have more than one, then press the round record button. The browser shows its microphone prompt, and the first time you allow it the level meter begins drawing your voice.
- Record, and pause instead of restarting. The timer counts up in minutes and seconds while you speak. Press the middle button to freeze both the capture and the clock, and press it again to carry on in the same file, so a stumble costs you nothing.
- Stop and download. The square button closes the take. The result panel shows the size and the length, and the download button saves the file with the extension that matches what your browser actually recorded.
About the Allow microphone button
The page asks for nothing when it loads. The button labelled Allow microphone exists for anyone who wants to grant access and check the input before committing to a take: it opens the prompt, then immediately releases the tracks and marks the microphone as ready, so nothing is being recorded while you test your levels. Pressing record does the same thing on its own, so you can ignore the extra button entirely if you prefer a single step.
What pause actually does
Pause is a real hold, not a restart. The browser stops writing audio to the file and the on-screen timer stops with it, so the duration you see at the end is the duration of the audio rather than the length of the session. Resume continues the same stream, which is why a ninety second recording with three pauses is still one file instead of four.
Which format you get, and why the extension changes
The recorder does not pick a container so much as accept the best one your browser offers. It asks, in order, for Opus inside WebM, then plain WebM, then Opus inside Ogg, and finally MP4, and it records with the first entry the browser confirms.
| Browser | What it usually records | Download extension |
|---|---|---|
| Chrome, Edge | WebM with Opus | .webm |
| Firefox | WebM with Opus, Ogg on builds that only offer that | .webm or .ogg |
| Safari | MP4 with AAC | .m4a |
The download button reads the format back out of the recorder rather than guessing from the browser name, so the file name always describes the file you actually have. That is a small detail with a practical payoff: you never end up with a file called .webm that is secretly an MP4, which is the kind of mismatch that makes an editor refuse a perfectly good recording.
Converting the take to MP3 in the browser
The recorder can hand you an MP3 without a server, using a WebAssembly build of FFmpeg that runs in the tab. The conversion writes a 192 kbps file that plays in any editor, phone or car stereo, which matters because not every destination accepts WebM or Ogg.
The catch is the engine itself. That build is around 30 megabytes, and the page does not fetch it in the background, so the first time you press Convert to MP3 you watch a progress bar fill while the encoder is downloaded. Later conversions in the same tab reuse it and start almost at once. If a raw recording is all you need, the encoder is never loaded at all.
Where a browser recorder stops
The trade for having nothing to install is a fixed set of things this page cannot do. The table below puts it beside the two alternatives most people reach for.
| Browser voice recorder | Desktop system recorder | Phone's built-in recorder | |
|---|---|---|---|
| Install required | No | Yes | Nothing extra |
| Where the audio goes | Stays on your device | Stays on your device | Stays on your device |
| Output container | Chosen by the browser: WebM, Ogg or MP4 | Usually WAV or MP3, selectable | M4A or MP4 |
| Quality settings | None | Bitrate and sample rate selectable | A few presets |
| Tracks per take | One | Multi-track possible | One |
| Editing in the same app | None | Trim and levels built in | Trim only |
Beyond that table, four limits come up again and again in practice.
- The container is not yours to choose. There is no format menu. Chrome and Edge usually write WebM with Opus, Safari writes MP4 with AAC, and the download extension follows the file you actually got, so the same take can arrive as .webm on one machine and .m4a on another.
- The first MP3 costs a wait. Nothing is downloaded until you ask for it, which keeps the page fast, but it also means the first conversion includes roughly 30 megabytes of encoder before any audio is processed.
- One track per take. The recorder holds a single microphone stream, so you cannot record a second voice or an instrument on its own track and mix them later.
- No monitoring while you record. The level meter shows your input as a moving waveform, but the audio is never routed to your speakers, so you cannot hear yourself live.
- Secure context only. Microphone access needs HTTPS, so a plain http address or a file opened from disk is refused before the page can ask.
Audio Trimmer
A take is one continuous file, so cut the dead air off the front or drop a mistake out of it after you download. Still on your device.
Does it work on a phone, and does it need HTTPS?
Yes to both. Microphone capture is limited to secure contexts, so the page has to be served over HTTPS; on a plain http address, or a file opened straight from disk, the browser refuses the request before the page can ask. This site is served over HTTPS end to end, so the prompt appears normally.
Mobile browsers expose the microphone to web pages as well, so the same recorder runs in Safari on an iPhone and in Chrome on Android, and the layout folds down to a single column. What the phone takes away is choice, because the browser decides which microphone to open, and a long take sits in memory, so short clips are comfortable and very long ones are not.
If you also need the picture, the companion guides cover recording from your webcam and recording your screen without installing anything. Both work the same way, in a tab, with the file staying on your device. This page is the audio-only one.