Convert MKV to OPUS Online
Extract audio from MKV video to modern Opus, free, in your browser.
.mkv · up to 100 MB
What it's for
MKV soundtracks to Opus for streaming and Discord bots
Stream copy when possible
If the MKV already contains Opus, FFmpeg performs direct extraction with no re-encoding. Zero additional loss.
AC-3, DTS, and AAC supported
Convert any audio codec from MKV to modern Opus with a single click.
100% private
Your MKV video never leaves your device. Fully local processing with FFmpeg.wasm.
Perfect for Discord bots
discord.js and discord.py bots play Opus directly, with no real-time transcoding needed.
How it works
Three steps, no hassle
Upload your MKV file
Drag or select your .mkv video. Works with movies, anime, and desktop recordings.
Smart stream extraction
If the MKV already contains Opus, FFmpeg performs a direct stream copy — no re-encoding. If it is AC-3, DTS, or AAC, it converts to Opus.
Download the audio
Get your .opus for Discord, streaming, bots, or archiving. Perfect for anime and movie soundtracks.
FAQ
Got questions?
The Matroska container (MKV, spec published in 2003 by Steve Lhomme's team at matroska.org) supports virtually any audio codec: AC-3 (Dolby Digital) is common in Blu-ray rips; DTS and DTS-HD in high-quality rips; AAC in anime encodes; MP3 in older rips; FLAC in lossless encodes; and Opus since approximately 2013, when FFmpeg and HandBrake added native support. Anime fansubbing groups largely migrated to Opus between 2016 and 2018 for its efficiency at 64–128 kbps.
Stream copy is the operation where FFmpeg extracts the audio stream from the MKV container without decoding or re-encoding it, simply repackaging it in OGG. This happens when the MKV already contains Opus audio. The result is bit-for-bit identical to the original audio: no generation loss. FFmpeg automatically detects the stream codec; if it is Opus, it performs -c:a copy. If it is AC-3, DTS, AAC, or any other codec, it decodes to PCM and re-encodes to Opus.
Technically yes, the conversion works with any MKV. Legal considerations are the user's responsibility: MKV files ripped from your own Blu-rays are legal in many jurisdictions for personal use; those downloaded from torrent sources may involve copyright infringement depending on local law. The tool does not discriminate on content and processes any valid .mkv file.
Yes. Anime MKVs often include the original Japanese audio (JPN), English dub (ENG), and sometimes director's commentary. Blu-ray rips may include DTS-HD MA (English), AC-3 (other languages), and Opus (streaming versions). By default, FFmpeg extracts the first audio track marked as default in the container. To select a specific track (e.g., the JPN track in a multi-language anime MKV), you would need to specify the stream index.
For anime or movie soundtracks with high dynamic range: 128–160 kbps offers perceptually transparent quality per MUSHRA studies. For voice dialogue with moderate sound effects: 96 kbps. For dialogue only (audiobooks, director's commentary): 48–64 kbps. If the original MKV already has Opus (stream copy), the output bitrate matches the original.
Yes. Discord bots built with discord.py, discord.js (using @discordjs/voice), JDA (Java Discord API), or Javacord can play Opus files directly in voice channels. The @discordjs/voice library accepts an Opus stream as input for OpusEncoder, making .opus files produced here directly compatible. The playback latency of a pre-encoded .opus file is lower than an MP3 file requiring real-time decoding and re-encoding.
Convert MKV to Opus: extract movie and anime soundtracks with intelligent stream copy
Matroska (MKV) is an open-source video container whose specification was first published in December 2003 by the matroska.org team led by Steve Lhomme. Unlike proprietary containers such as MOV (Apple) or AVI (Microsoft), Matroska was designed from the start to support an unlimited number of video, audio, subtitle, and metadata tracks in a single file, with no codec restrictions. This flexibility made it the preferred format for the digital video distribution community: anime fansubbing groups, Blu-ray rip distributors, and HandBrake and MakeMKV users for personal collection ripping. Audio inside an MKV can be any codec: AC-3 (Dolby Digital, ATSC A/52) in Region 1 Blu-ray rips; DTS or DTS-HD Master Audio in high-fidelity rips; AAC-LC or HE-AAC in anime encodes optimized for distribution; FLAC in audiophile lossless encodes; MP3 in pre-2010 DVD rips; and Opus since approximately 2013, when updated versions of FFmpeg, HandBrake, and MKVToolNix added full Opus-in-Matroska write support. The fansubbing group Erai-Raws, one of the largest for anime simulcast releases, began using Opus by default in their releases in 2017; other large groups such as SubsPlease and Judas adopted it between 2018 and 2020. The need to extract audio from MKV to Opus comes from multiple sources: Discord bots serving soundtrack audio in voice channels, archiving movie audio with a modern codec, and extracting director's commentaries for independent listening.
The most technically relevant feature of this conversion is the stream copy capability when the MKV already contains Opus audio. Stream copy is the operation where FFmpeg extracts the audio stream from the Matroska container without any decoding or re-encoding process, repackaging it directly into an OGG container. The equivalent FFmpeg command-line operation is ffmpeg -i input.mkv -vn -acodec copy output.opus. The result is bit-perfect relative to the original Opus: no additional generation loss, no modification of samples, and processing speed is limited only by file read speed, not CPU. This is especially relevant for anime MKV files distributed by groups like Erai-Raws and SubsPlease that already include Opus: the stream copy produces a .opus that is literally the same audio that was in the MKV. For MKV files containing AC-3 (Dolby Digital at 640 kbps in North American Blu-ray), DTS (1536 kbps in high-fidelity Blu-ray), or AAC, the conversion introduces one transcoding generation: the audio is decoded to PCM and re-encoded to Opus. In this case, 128–160 kbps is recommended for musical content with high dynamic range (orchestral soundtracks, action sound effects) and 96 kbps for predominantly dialogue content. Xiph.org's 2012 MUSHRA studies established that Opus at 128 kbps is perceptually equivalent to AAC at 192 kbps on complex musical material.
Convertir.ai executes the MKV audio extraction to Opus entirely in the browser with FFmpeg.wasm, without sending any video data to external servers. The process: FFmpeg.wasm reads the MKV, analyzes the available streams, identifies the primary audio stream (codec, sample rate, channel count, bitrate), and determines whether to apply stream copy or transcoding. If the audio stream is Opus, it applies -vn -c:a copy: extraction takes only the seconds needed to read the container with no CPU encoding load. If the stream is AC-3, DTS, AAC, MP3, or other, it decodes to PCM at 48 kHz (converting non-standard sample rates with FFmpeg's Kaiser sinc filter) and encodes with libopus at the chosen bitrate. For multichannel AC-3 (5.1, 7.1), audio is downmixed to stereo using FFmpeg's default downmix matrix (Lt/Rt downmix) before Opus encoding, as Opus in OGG supports up to 8 channels but most players and platforms expect stereo or mono. Text metadata from the MKV (title, artist, album, date in the Matroska segment info) is transferred to Opus Vorbis Comment tags. The output .opus file in OGG container is compatible with VLC 2.1+, mpv, Firefox 15+, Chrome 25+, Foobar2000, all Discord clients, and the bot frameworks discord.js (@discordjs/voice), discord.py (using PyNaCl), JDA, and Javacord. The service is completely free, no signup, and no file limit.