Initializing... drag & drop files here
Supports: WEBM
.webm videos. Batch is supported — queue dozens at once.WebM is a container, not a codec. A .webm file from YouTube, OBS, Chrome's MediaRecorder API, or a Discord screen capture usually carries VP8/VP9 video plus an Opus or Vorbis audio track. Extracting that audio to a standalone .opus file is the cleanest way to keep just the sound — for a podcast clip, a voice memo, a music sample, or a transcription pipeline — at the smallest size today's codecs can deliver.
| Property | WebM (Opus track) | .opus File |
|---|---|---|
| Container | Matroska-derived (Google) | Ogg |
| Video stream | VP8 / VP9 / AV1 + audio | Audio only |
| Audio codec | Opus or Vorbis | Opus only |
| Typical size (10-min screen rec) | 50–500 MB | 5–15 MB |
| Re-encode required? | No, if source is Opus (stream copy) | No, if source is Opus |
| Best for | Web video, MediaRecorder, YouTube DASH | Voice memos, podcasts, audio-only delivery |
| Native player support | Chrome, Firefox, Edge, Safari 14.1+ | VLC, foobar2000, Chrome, Firefox, Safari 11+ |
| Use case | Bitrate | Channel | Sample Rate | Notes |
|---|---|---|---|---|
| Voice memo / narrow speech | 16–32 kbps | Mono | 16000 Hz | SILK layer; intelligible down to 6 kbps |
| Podcast / interview | 32–64 kbps | Mono | 24000 Hz | Comparable to MP3 at 96 kbps |
| Stereo music (casual) | 96 kbps | Stereo | 48000 Hz | Transparent for most listeners |
| Stereo music (audiophile) | 128–160 kbps | Stereo | 48000 Hz | Xiph considers ~128 kbps perceptually transparent |
| Archival / max quality | 256–510 kbps | Stereo | 48000 Hz | Diminishing returns past 256 kbps |
If you pick the Highest preset and leave Audio Channel and Sample Rate at Original, the source Opus stream is muxed straight into the Ogg container without re-encoding (a stream copy). That's lossless and finishes in a second or two even for long files. Changing the bitrate, sample rate, or channel count forces a re-encode.
Open the file in VLC and check Tools → Codec Information, or run ffprobe file.webm from the command line. YouTube downloads since late 2014, Chrome's MediaRecorder API, OBS WebM output, and Discord screen recordings all use Opus by default. Older WebM files (pre-2013) and some legacy YouTube uploads still use Vorbis — those will always require a re-encode here.
Opus is a hybrid codec that switches between two engines: SILK (from Skype, tuned for speech) and CELT (from Xiph, tuned for music), often blending them per frame. It was designed in 2010–2012 with 20+ years of psychoacoustic research baked in, while MP3 dates to 1993. Independent listening tests (Hydrogenaudio, Xiph) consistently show Opus at 64 kbps matching MP3 at 128 kbps, and Opus at 96 kbps matching MP3 at 160–192 kbps.
.opus files?Native Opus support landed in iOS 11 and macOS High Sierra (Sept 2017). Files Live Player, VLC for iOS, and the Files app all play .opus directly. On older iOS, install VLC. Note: AirDrop and the macOS Finder Quick Look preview work; iTunes / Apple Music library import does not — convert to AAC for the iTunes library.
There's no hard cap that you'll hit for typical use — multi-GB files are fine because processing runs in your browser tab against our edge nodes, not a single small server slice. For multi-hour 4K screen recordings, expect the upload to be the slow part; the conversion itself is fast because Opus extraction from a WebM that already contains Opus is just a remux.
VBR is the right default and what the reference opusenc uses unless you override it. It allocates more bits to complex passages (music, multiple voices) and fewer to silence, giving better quality per byte. Pick CBR only when the downstream expects a predictable rate — some streaming endpoints, hardware decoders, or fixed-bandwidth voice channels.
Yes. In Advanced Options expand Trim, enter start 00:12:00.000 and duration 00:03:00.000 in the HH:MM:SS.ms boxes. The output is a 3-minute .opus cut. If you want lossless trimming on a clip boundary, keep all other settings at Original so the operation stays a stream copy.
Whisper, Deepgram, Google Speech-to-Text, and Azure Speech all accept Opus natively and internally resample to 16 kHz mono anyway. A 1-hour interview is roughly 600 MB as 48 kHz/16-bit WAV but only 14 MB as 32 kbps Opus mono — same recognition accuracy, 40× smaller upload. For long-form, convert WebM to MP3 only if your tool predates 2018 and lacks Opus support.
This page targets the Opus container specifically. If you need MP3 instead (for older car stereos, iTunes import, or legacy tools), use WebM to MP3. For lossless WAV (audio editing, mastering), use WebM to WAV. For AAC inside an M4A (Apple ecosystem), use WebM to M4A.