Voice recorder

Voice Recorder Online

QUICK ANSWER

Record voice online free, right in your browser. One big mic button, a live level meter so you know your input is hot, and a one-click export to MP3. Voice memos, podcast scratch tracks, language-learning practice, dictation, lecture notes — captured without installing a thing. Nothing is uploaded; the recording stays in your tab until you save it.

  • Runs in your browser
  • Nothing uploaded
  • No signup
  • Free forever
  • MP3 export
  • Live level meter
00:00
Tap the mic to start
Input level0%

Recording stays on your device. The MP3 export runs FFmpeg in your browser — no upload. Have a video file instead?

Trusted by creators at startups, agencies, and Fortune 500 teams.
Free forever — no signup, no watermark, no length cap.

How it works

  1. 1

    Allow microphone access and click record

    On first visit the browser shows a one-time permission prompt — click Allow. The big mic button starts a getUserMedia stream feeding both a MediaRecorder (capturing WebM/Opus) and a Web Audio AnalyserNode (driving the live level meter). Speak normally and watch the meter respond to your voice in real time.

  2. 2

    Stop when you are done and preview the take

    Click Stop to finalize the recording. An audio player appears so you can scrub through, listen back, and decide whether to keep the take or start over. The whole take lives in tab memory — closing the tab discards it unless you download.

  3. 3

    Download as WebM (instant) or export as MP3

    WebM/Opus is the browser's native output and downloads instantly. For universal compatibility — podcast hosts, transcript services, email attachments — click Export MP3 and we re-encode to MP3 with FFmpeg in your browser via WebAssembly. First MP3 export downloads the FFmpeg core (~25 MB) once; subsequent exports are fast.

When you should use a real DAW instead

This recorder is deliberately a single-track scratchpad. The moment you need multi-track recording, noise gating, EQ, compression, or music layered under the voice, reach for Audacity, GarageBand, Reaper, or a hosted tool like Riverside or Descript. What this page is great at: turning a fleeting thought into an audio file in under ten seconds, no install, no signup.

Recording your screen alongside the voice?

For voice plus screen capture with a shareable link at the end, that is exactly what Clipy does — same no-signup, no-watermark policy, plus a viewer page anyone can open. Need to verify mic and camera before a Zoom or Google Meet call? Zoom mic and camera test, Google Meet mic and camera test, and Microsoft Teams mic and camera test cover the meeting-platform-specific gotchas in 30 seconds each.

What this is for

Voice memos for a podcast intake, dictation you will feed into a transcript, a quick audio note for a teammate, scratch tracks while you sketch out a song idea, a language-learning student listening back to their own pronunciation, a lecturer leaving an audio annotation for students — anything where you need to capture your voice without setting up a DAW. Open the page, hit the mic, talk, stop, download. That is the whole product. If you are recording a podcast for distribution, multitracking guests, applying compression or layering in music, you want Audacity, GarageBand, Riverside, or Descript. This tool is intentionally a single-track scratchpad — minimal, fast, no signup gate.

Opus by default, MP3 on demand

Native browser recording goes to audio/webm;codecs=opus. Opus is the right call for storage and quality: it sounds excellent at low bitrates (64–128 kbps for voice is typically indistinguishable from uncompressed) and is what most modern tools and APIs accept. The WebM download happens instantly because no re-encoding is needed. MP3 is still the lingua franca for podcast hosts, transcript services, voicemail systems, and old-school email attachments. Our MP3 export runs FFmpeg in your browser via WebAssembly with libmp3lame -q:a 2 (VBR around 190 kbps) — broadcast quality for voice with no audible artifacts. Same engine as our MP4 → MP3 extractor, just different input.

The level meter is doing real work

We tap the live mic stream into a Web Audio AnalyserNode and compute peak amplitude every animation frame. The bar going red means you are clipping — back off the mic or lower system input gain. The bar staying flat means the OS is feeding us the wrong device or the mic is muted at the hardware level. It is the feedback most browser recorders skip, and it is the difference between a usable take and a wasted one you only notice on playback.

Common questions

Is the recording uploaded anywhere?

No. The mic stream goes through MediaRecorder in your browser tab, the resulting blob lives in tab memory, and the MP3 export runs in-browser with FFmpeg/WebAssembly. Nothing leaves your device unless you upload the downloaded file yourself. Open the browser network tab while recording — zero requests go out.

Why does my MP3 take a few seconds longer than the WebM?

Because MP3 export re-encodes the audio. WebM/Opus is what the browser already produced — it is instant. MP3 needs to load FFmpeg (~25 MB the first time, cached after) and re-encode through libmp3lame. After the first run FFmpeg stays warm and subsequent exports are quick.

What is the bitrate and quality?

Native: Opus at whatever the browser picks (typically 64–128 kbps for voice — already excellent). MP3 export: VBR around 190 kbps via -q:a 2, which is broadcast-quality for voice and leaves no audible artifacts even on close listening.

My level meter is completely flat. What is wrong?

Four common causes. (1) Wrong input device selected — pick the right one from the dropdown. (2) Mic muted at the OS level (Windows: Settings > System > Sound > Input; macOS: System Settings > Sound > Input). (3) Physical mute switch on a headset cable or earcup. (4) Browser denied permission silently — refresh the tab and click Allow when prompted, or click the lock icon in the address bar and grant microphone access.

Can I record for an hour or longer?

Yes, but the blob lives in your tab memory, so very long recordings can get sluggish on lower-end machines. For anything past ~30 minutes a desktop recorder is more comfortable. This tool is built for sub-30-minute clips — voice memos, scratch takes, lecture segments.

Why does my MP3 sound different from the WebM?

Both are lossy encodings but they use different algorithms. At the bitrates we use, the difference is imperceptible on voice content — A/B comparison on speech will not reveal which is which. If you do hear a difference, it is almost always slight high-frequency softening from MP3, which is harmless for spoken-word use cases.

Does the voice recorder work on iPhone and Android?

Yes on modern Safari (iOS 14.1+) and Chrome (Android). MediaRecorder is supported, the level meter works through Web Audio, and the MP3 export runs the same FFmpeg WebAssembly module — though older mobile devices may hit memory ceilings on longer recordings.

Why a browser voice recorder instead of a native app?

Speed and trust. There is nothing to install, no permission grab beyond the one mic prompt, and you can see in the network tab that nothing leaves your device. For a single-track scratch recording, opening a tab is faster than launching any DAW.

More free tools