Single-model take on phase correction. MelBand Roformer separates the track; a spectral invert rebuilds the instrumental as mixture minus vocals (carrying the mixture phase), and the result keeps the model instrumental magnitude while borrowing that invert phase. One model, no second separation pass.
Monthly usage: 6 997, Monthly rating: 4.5 (39 votes)Weighted-average ensemble of three models β BS Roformer 124-bands (counted twice), SCNet, and MelBand Roformer becruily deux β producing BOTH a vocals and an instrumental stem. The strongest model leads each ensemble (2x), while the other two temper its artefacts. One separation pass per model feeds both outputs.
Monthly usage: 1 687, Monthly rating: 4.9 (9 votes)Keeps the spectrum of the HyperACE v2 instrumental model but resynthesises it on the phase of the 124 bands model, whose phase is cleaner. Removes phase smearing without giving up the fuller HyperACE timbre.
Monthly usage: 438, Monthly rating: 4.7 (6 votes)Extracts the vocal first, then splits it into lead and backing. Separating a clean vocal splits far better than running the karaoke model on the whole mix.
Monthly usage: 371, Monthly rating: 5 (3 votes)MVSep Karaoke: strips the lead vocal but keeps backing vocals with the instrumental β plus the isolated lead vocal as a second file.
Monthly usage: 3 239, Monthly rating: 4.6 (8 votes)Rebuilds the instrumental as original-minus-vocals: the extracted vocal is polarity-inverted and mixed back with the source, so the result keeps the original mixture's phase. Cleaner transients than a model's direct instrumental output.
Monthly usage: 6 548, Monthly rating: 4 (26 votes)Extracts the vocal, then removes room reverb from it for a dry, close acapella.
Monthly usage: 168, Monthly rating: 3.7 (3 votes)De-noises the track first, then extracts the vocal from the cleaned audio.
Monthly usage: 1 159, Monthly rating: 3.6 (8 votes)Removes the backing vocals and keeps the lead vocal with the instrumental β the reverse of the karaoke pack.
Monthly usage: 490, Monthly rating: 5 (3 votes)Pulls the drums out of the mix, then splits the kit into kick, snare, toms, hi-hat, ride and crash.
Monthly usage: 115, Monthly rating: 0 (0 votes)Removes both lead and backing vocals, keeping just the instrumental to sing over.
Monthly usage: 691, Monthly rating: 4.7 (6 votes)Ensemble of best vocal models. Algorithm gives the highest possible quality for vocal and instrumental stems. The latest ensemble consists of BS Roformer, MelBand Roformer and SCNet XL IHF vocal models.
Vocals Monthly usage: 17 363, Monthly rating: 3.5000 (24 votes)This ensemble is based on algorithm which took 2nd place at Music Demixing Track of Sound Demixing Challenge 2023. The main changes comparing to contest version is much better individual stem models.
Vocals Drums Bass Monthly usage: 2 149, Monthly rating: 4.6000 (5 votes)It's Ensemble (vocals, instrum, bass, drums, other) + more models included like guitars, piano, wind, strings, back/lead vocals and drumsep.
Featured Vocals Drums Bass Piano Guitar Wind Strings Keys Monthly usage: 4 529, Monthly rating: 3.2857 (14 votes)BS Roformer SW model, which generates 6 stems at once with superior quality.
Featured Vocals Drums Bass Piano Guitar Keys Monthly usage: 237 839, Monthly rating: 4.6997 (899 votes)Algorithm Demucs4 HT. It's fast and gives relatively good quality for bass/drums/other stems.
Vocals Drums Bass Monthly usage: 11 350, Monthly rating: 4.6667 (33 votes)BS Roformer model. Excellent quality for vocals/instrumental separation.
Featured Vocals Monthly usage: 145 125, Monthly rating: 4.5219 (274 votes)BS PolarFormer is a high-quality algorithm for vocal separation.
Monthly usage: 13 109, Monthly rating: 4.6667 (15 votes)Algorithm for separating tracks into vocal and instrumental parts based on the MelBand Roformer neural network
Vocals Monthly usage: 69 006, Monthly rating: 4.4408 (152 votes)Set of MDX23C models which is based on code released by kuielab for Sound Demixing Challenge 2023. Very good for vocals/instrumental separation.
Vocals Monthly usage: 5 430, Monthly rating: 4.5000 (14 votes)Algorithm for separating tracks into vocal and instrumental parts based on the SCNet neural network
Vocals Monthly usage: 4 288, Monthly rating: 4.5833 (12 votes)MDX B models are based on kuielab code from Music Demixing Challenge 2021. Models were retrained by UVR team on big dataset. For long time models were best for vocals/instrumental separation.
Vocals Monthly usage: 1 577, Monthly rating: 3.8571 (7 votes)A set of models from the Ultimate Vocal Remover program, which are based on the old VR architecture. Most of the models are vocal, but there are also special models for karaoke, piano, removing reverberation effects, etc.
Vocals Monthly usage: 7 054, Monthly rating: 3.0833 (12 votes)Demucs4 Vocals 2023 model - it's Demucs4 HT model fine-tuned on big vocals dataset.
Vocals Monthly usage: 1 065, Monthly rating: 2.0000 (1 votes)Algorithm for extracting only lead vocals and everything else based on the MelBand Roformer and SCNet models.
Vocals Monthly usage: 102 745, Monthly rating: 4.7407 (324 votes)The MDX-B Karaoke model was prepared as part of the Ultimate Vocal Remover project. The model produces high-quality lead vocal extraction from a music track.
Vocals Monthly usage: 7 168, Monthly rating: 4.2667 (15 votes)An unique model for removing crowd sounds from music recordings (applause, clapping, whistling, noise, laugh etc.).
Monthly usage: 11 906, Monthly rating: 4.1628 (43 votes)Medley Vox is an algorithm for separating multiple singers within a single music track and evaluation dataset for this task.
Vocals Monthly usage: 8 014, Monthly rating: 3.4545 (22 votes)MVSep Multichannel BS - uses the best vocal model to extract sound from multi-channel audio (5.1, 7.1, etc.).
Vocals Monthly usage: 2 694, Monthly rating: 4.3333 (3 votes)A model for separating male and female voices within a single vocal track. The track should contain only voices, no music.
Vocals Monthly usage: 11 239, Monthly rating: 2.9800 (50 votes)Choir Extraction Model
Vocals Monthly usage: 2 395, Monthly rating: 4.5556 (9 votes)Model to separate vocals and strings to SATB parts (Soprano, Alto, Tenor, and Bass)
Vocals Monthly usage: 13 650, Monthly rating: 4.4737 (19 votes)The MVSep Drums model produces high-quality separation of music into a drums part and everything else.
Drums Monthly usage: 19 979, Monthly rating: 4.5385 (52 votes)The MVSep Bass model produces high-quality separation of music into a bass part and everything else.
Bass Monthly usage: 12 284, Monthly rating: 4.5278 (36 votes)Synth extraction model
Monthly usage: 9 018, Monthly rating: 4.7869 (61 votes)The DrumSep model divides the drum track into several types: 'kick', 'snare', 'toms', 'cymbals' (it includes 'hh', 'ride', 'crash').
Drums Monthly usage: 22 182, Monthly rating: 4.9024 (379 votes)MVSep Piano model is based on MDX23C, MelRoformer and SCNet Large architectures. It produces high quality separation for piano and other stems.
Piano Keys Monthly usage: 12 574, Monthly rating: 4.1837 (49 votes)No data found
Piano Keys Monthly usage: 1 946, Monthly rating: 4.8750 (64 votes)The MVSep Keys is a high quality model for separating music into keys instruments and everything else.
Keys Monthly usage: 3 677, Monthly rating: 4.8857 (35 votes)The MVSep Organ model produces high-quality separation of music into an organ part and everything else.
Keys Monthly usage: 3 192, Monthly rating: 4.6290 (62 votes)No data found
Keys Monthly usage: 861, Monthly rating: 4.9091 (11 votes)No data found
Keys Monthly usage: 1 359, Monthly rating: 5.0000 (11 votes)No data found
Keys Monthly usage: 714, Monthly rating: 4.8070 (57 votes)No data found
Keys Monthly usage: 317, Monthly rating: 4.1250 (8 votes)The Rhodes (Fender Rhodes) is an iconic electromechanical keyboard instrument designed by Harold Rhodes.
Keys Monthly usage: 900, Monthly rating: 4.7600 (50 votes)The Clavinet is an electromechanical keyboard instrument developed and manufactured by the German company Hohner from 1964 to 1982.
Keys Monthly usage: 756, Monthly rating: 4.7308 (52 votes)Mellotron is a electromechanical keyboard instrument developed in England in the early 1960s.
Keys Monthly usage: 697, Monthly rating: 4.6667 (57 votes)The MVSep Guitar model produces high-quality separation of music into a guitar part (including acoustic and electronic) and everything else.
Guitar Strings Monthly usage: 9 648, Monthly rating: 4.1250 (16 votes)No data found
Guitar Strings Monthly usage: 5 458, Monthly rating: 4.8721 (86 votes)No data found
Guitar Strings Monthly usage: 5 542, Monthly rating: 4.8810 (42 votes)No data found
Guitar Strings Monthly usage: 15 535, Monthly rating: 4.4390 (82 votes)Algorithm for isolating pedal steel guitar sounds from a melody
Guitar Monthly usage: 758, Monthly rating: 5.0000 (12 votes)The MVSep Plucked Strings is a high quality model for separating music into plucked strings instruments and everything else.
Strings Monthly usage: 1 779, Monthly rating: 4.8333 (6 votes)No data found
Strings Monthly usage: 824, Monthly rating: 5.0000 (6 votes)No data found
Strings Monthly usage: 371, Monthly rating: 5.0000 (1 votes)No data found
Strings Monthly usage: 535, Monthly rating: 4.8000 (5 votes)No data found
Strings Monthly usage: 286, Monthly rating: 5.0000 (2 votes)No data found
Strings Monthly usage: 275, Monthly rating: 4.0000 (1 votes)No data found
Strings Monthly usage: 366, Monthly rating: 4.7500 (8 votes)The MVSep Bowed Strings is a high quality model for separating music into bowed string instruments and everything else.
Strings Monthly usage: 9 980, Monthly rating: 4.8843 (121 votes)No data found
Strings Monthly usage: 3 184, Monthly rating: 4.8387 (31 votes)No data found
Strings Monthly usage: 825, Monthly rating: 4.8571 (7 votes)No data found
Strings Monthly usage: 1 271, Monthly rating: 5.0000 (15 votes)No data found
Strings Monthly usage: 1 215, Monthly rating: 4.3333 (3 votes)The MVSep Wind model produces high-quality separation of music into a wind part and everything else.
Wind Monthly usage: 5 862, Monthly rating: 4.6500 (20 votes)The MVSep Brass is a high quality model for separating music into brass wind instruments and everything else.
Wind Monthly usage: 3 793, Monthly rating: 4.5849 (53 votes)The MVSep Woodwind is a high quality model for separating music into woodwind instruments and everything else.
Wind Monthly usage: 1 432, Monthly rating: 4.6667 (9 votes)No data found
Wind Monthly usage: 3 836, Monthly rating: 4.6250 (48 votes)No data found
Wind Monthly usage: 2 188, Monthly rating: 4.6667 (18 votes)No data found
Wind Monthly usage: 2 067, Monthly rating: 4.7500 (32 votes)No data found
Wind Monthly usage: 974, Monthly rating: 5.0000 (16 votes)No data found
Wind Monthly usage: 425, Monthly rating: 4.6667 (3 votes)No data found
Wind Monthly usage: 757, Monthly rating: 5.0000 (3 votes)No data found
Wind Monthly usage: 665, Monthly rating: 4.9474 (38 votes)No data found
Wind Monthly usage: 841, Monthly rating: 5.0000 (19 votes)No data found
Wind Monthly usage: 479, Monthly rating: 0 (0 votes)No data found
Wind Monthly usage: 375, Monthly rating: 4.9231 (26 votes)The MVSep Bagpipes model provides high-quality music separation, isolating the bagpipes from the rest of the track.
Wind Monthly usage: 268, Monthly rating: 4.4000 (30 votes)No data found
Wind Monthly usage: 697, Monthly rating: 4.1875 (16 votes)The MVSep Percussion is a high quality model for separating music into percussion instruments and everything else.
Percussion Monthly usage: 4 192, Monthly rating: 4.3929 (56 votes)No data found
Percussion Monthly usage: 971, Monthly rating: 4.9474 (19 votes)No data found
Percussion Monthly usage: 534, Monthly rating: 0 (0 votes)No data found
Percussion Monthly usage: 548, Monthly rating: 4.6000 (5 votes)No data found
Percussion Monthly usage: 681, Monthly rating: 5.0000 (1 votes)No data found
Percussion Monthly usage: 234, Monthly rating: 4.7500 (4 votes)No data found
Percussion Monthly usage: 1 150, Monthly rating: 5.0000 (16 votes)No data found
Percussion Monthly usage: 1 548, Monthly rating: 4.8333 (36 votes)No data found
Percussion Monthly usage: 332, Monthly rating: 5.0000 (2 votes)No data found
Percussion Monthly usage: 643, Monthly rating: 5.0000 (5 votes)No data found
Percussion Monthly usage: 453, Monthly rating: 0 (0 votes)No data found
Percussion Monthly usage: 1 096, Monthly rating: 4.9231 (13 votes)No data found
Percussion Monthly usage: 544, Monthly rating: 4.8182 (11 votes)No data found
Monthly usage: 2 919, Monthly rating: 3.7333 (15 votes)BandIt Plus model for separating tracks into speech, music and effects.
Monthly usage: 1 949, Monthly rating: 3.0000 (2 votes)Bandit v2 is a model for cinematic audio source separation in 3 stems: speech, music, effects/sfx. It was trained on DnR v3 dataset.
Monthly usage: 2 470, Monthly rating: 2.9167 (12 votes)MVSep DnR v3 is a cinematic model for splitting tracks into 3 stems: music, sfx and speech.
Monthly usage: 19 438, Monthly rating: 3.2941 (17 votes)The MVSep Braam model provides high-quality music separation, isolating the cinematic "Braam" sound effect from the rest of the track.
Monthly usage: 331, Monthly rating: 4.8000 (5 votes)Algorithm for isolating "Riser" sounds from a musical melody
Monthly usage: 440, Monthly rating: 4.3333 (9 votes)No data found
Monthly usage: 9 829, Monthly rating: 4.6182 (55 votes)The algorithm restores the quality of audio. For example MP3 files compressed to 128 kbps or lower and other types.
Super Resolution Monthly usage: 9 708, Monthly rating: 3.1500 (20 votes)Set of different models to remove reverberation effect from music.
Monthly usage: 15 585, Monthly rating: 3.6905 (42 votes)No data found
Monthly usage: 21 621, Monthly rating: 4.5776 (116 votes)Algorithm AudioSR: Versatile Audio Super-resolution at Scale. Algorithm restores high frequencies.
Super Resolution Monthly usage: 5 896, Monthly rating: 3.7105 (38 votes)FlashSR - audio super resolution algorithm for restoring high frequencies
Super Resolution Monthly usage: 2 871, Monthly rating: 3.3333 (12 votes)Generating audio based on a given text prompt
Monthly usage: 478, Monthly rating: 4.9680 (125 votes)Whisper is a pre-trained model for automatic speech recognition (ASR) and speech translation.
Monthly usage: 936, Monthly rating: 2.8750 (8 votes)Parakeet by NVIDIA is a state-of-the-art automatic speech recognition (ASR) model designed for accurate and efficient conversion of spoken language into text.
Monthly usage: 195, Monthly rating: 1.0000 (1 votes)VibeVoice - is a model for generating natural conversational dialogues from text with the ability to use a reference voice for cloning purposes.
Monthly usage: 3 695, Monthly rating: 4.0000 (10 votes)VibeVoice (TTS) - is a model for generating natural conversational dialogues from text, capable of creating dialogues with up to 4 speakers and durations of up to 90 minutes.
Monthly usage: 43, Monthly rating: 2.5000 (2 votes)Qwen3-TTS (Custom Voice) is a powerful speech generation model with predefined speakers
Monthly usage: 22, Monthly rating: 0 (0 votes)Qwen3-TTS (Voice Design) is a powerful speech generation model capable of creating a voice based on a description.
Monthly usage: 71, Monthly rating: 0 (0 votes)Qwen3-TTS (Voice Cloning) is a powerful speech generation model capable of cloning a voice based on a reference audio file
Monthly usage: 410, Monthly rating: 3.0000 (2 votes)A model for separating a musical composition into 53 different musical instruments
Vocals Drums Bass Piano Guitar Wind Strings Percussion Keys Monthly usage: 9 834, Monthly rating: 4.0000 (27 votes)Advanced AI Model for Stereo-to-Surround Upmixing
Monthly usage: 1 862, Monthly rating: 3.9143 (35 votes)A model that recovers pure dry lead vocals from processed audio
Monthly usage: 11 342, Monthly rating: 4.3898 (59 votes)Bark - is a generative model for realistic speech, music, and emotions.
Monthly usage: 26, Monthly rating: 1.0000 (1 votes)MVSep MultiSpeaker (MDX23C) - this model tries to isolate the most loud voice from all other voices.
Monthly usage: 1 015, Monthly rating: 3.0000 (3 votes)UniverSR - audio super resolution algorithm for restoring high frequencies
Monthly usage: 879, Monthly rating: 2.5758 (66 votes)Π‘lean vocals (acapellas) from artifacts after phase inversion
Monthly usage: 412, Monthly rating: 4.7143 (14 votes)The algorithm adds "whispering" effect to vocals.
Monthly usage: 777, Monthly rating: 4.5000 (2 votes)Algorithm for extracting the phantom center from a audio track
Monthly usage: 2 761, Monthly rating: 4.4286 (7 votes)Matchering is a novel tool for audio matching and mastering.
Monthly usage: 3 578, Monthly rating: 5.0000 (9 votes)SOME is a MIDI extractor that can convert singing voice to MIDI sequence.
MIDI Monthly usage: 628, Monthly rating: 1.0000 (1 votes)High-quality transcription of piano music into MIDI
MIDI Monthly usage: 2 387, Monthly rating: 3.7000 (10 votes)Basic Pitch is a modern neural network that converts melodic audio recordings into notes (MIDI format).
MIDI Monthly usage: 1 468, Monthly rating: 3.0000 (4 votes)Automatic transcription of drum audio to MIDI
MIDI Monthly usage: 916, Monthly rating: 3.0000 (2 votes)Multi-instrument MIDI transcription
MIDI Monthly usage: 56, Monthly rating: 0 (0 votes)HeartMuLa - an open-source AI model for music generation (an alternative to Suno and Udio)
Monthly usage: 147, Monthly rating: 3.0000 (2 votes)Algorithm Demucs3 (A and B versions)
Vocals Drums Bass Monthly usage: 227, Monthly rating: 0 (0 votes)No data found
Vocals