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.
Oylik foydalanish: 13 240, Oylik reyting: 4.4 (53 ovoz)Weighted-average ensemble of three models — BS Roformer 124-bands (counted twice), SCNet, and MelBand Roformer becruily deux — 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.
Oylik foydalanish: 2 309, Oylik reyting: 3.9 (12 ovoz)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.
Oylik foydalanish: 807, Oylik reyting: 5 (3 ovoz)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.
Oylik foydalanish: 551, Oylik reyting: 4.8 (4 ovoz)MVSep Karaoke: strips the lead vocal but keeps backing vocals with the instrumental — plus the isolated lead vocal as a second file.
Oylik foydalanish: 3 370, Oylik reyting: 4.6 (13 ovoz)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.
Oylik foydalanish: 1 974, Oylik reyting: 3.9 (13 ovoz)Extracts the vocal, then removes room reverb from it for a dry, close acapella.
Oylik foydalanish: 267, Oylik reyting: 1 (1 ovoz)De-noises the track first, then extracts the vocal from the cleaned audio.
Oylik foydalanish: 626, Oylik reyting: 5 (2 ovoz)Removes the backing vocals and keeps the lead vocal with the instrumental — the reverse of the karaoke pack.
Oylik foydalanish: 443, Oylik reyting: 4.5 (2 ovoz)Pulls the drums out of the mix, then splits the kit into kick, snare, toms, hi-hat, ride and crash.
Oylik foydalanish: 227, Oylik reyting: 0 (0 ovoz)Removes both lead and backing vocals, keeping just the instrumental to sing over.
Oylik foydalanish: 539, Oylik reyting: 0 (0 ovoz)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.
Vokal Oylik foydalanish: 12 665, Oylik reyting: 4.1905 (21 ovoz)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.
Vokal Barabanlar Bas Oylik foydalanish: 2 319, Oylik reyting: 3.6000 (5 ovoz)It's Ensemble (vocals, instrum, bass, drums, other) + more models included like guitars, piano, wind, strings, back/lead vocals and drumsep.
Tavsiya etilgan Vokal Barabanlar Bas Pianino Gitara Nafasli asboblar Torli asboblar Klavishli asboblar Oylik foydalanish: 5 563, Oylik reyting: 3.6364 (11 ovoz)BS Roformer SW model, which generates 6 stems at once with superior quality.
Tavsiya etilgan Vokal Barabanlar Bas Pianino Gitara Klavishli asboblar Oylik foydalanish: 267 947, Oylik reyting: 4.6615 (1031 ovoz)Algorithm Demucs4 HT. It's fast and gives relatively good quality for bass/drums/other stems.
Vokal Barabanlar Bas Oylik foydalanish: 10 943, Oylik reyting: 4.6471 (51 ovoz)BS Roformer model. Excellent quality for vocals/instrumental separation.
Tavsiya etilgan Vokal Oylik foydalanish: 160 390, Oylik reyting: 4.4460 (213 ovoz)BS PolarFormer is a high-quality algorithm for vocal separation.
Oylik foydalanish: 11 858, Oylik reyting: 4.5429 (35 ovoz)Algorithm for separating tracks into vocal and instrumental parts based on the MelBand Roformer neural network
Vokal Oylik foydalanish: 65 199, Oylik reyting: 4.1375 (80 ovoz)Set of MDX23C models which is based on code released by kuielab for Sound Demixing Challenge 2023. Very good for vocals/instrumental separation.
Vokal Oylik foydalanish: 5 693, Oylik reyting: 4.1500 (20 ovoz)Algorithm for separating tracks into vocal and instrumental parts based on the SCNet neural network
Vokal Oylik foydalanish: 4 990, Oylik reyting: 4.4000 (10 ovoz)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.
Vokal Oylik foydalanish: 1 614, Oylik reyting: 3.5000 (2 ovoz)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.
Vokal Oylik foydalanish: 6 980, Oylik reyting: 5.0000 (4 ovoz)Demucs4 Vocals 2023 model - it's Demucs4 HT model fine-tuned on big vocals dataset.
Vokal Oylik foydalanish: 933, Oylik reyting: 5.0000 (1 ovoz)Algorithm for extracting only lead vocals and everything else based on the MelBand Roformer and SCNet models.
Vokal Oylik foydalanish: 108 090, Oylik reyting: 4.5871 (402 ovoz)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.
Vokal Oylik foydalanish: 6 814, Oylik reyting: 4.3636 (11 ovoz)An unique model for removing crowd sounds from music recordings (applause, clapping, whistling, noise, laugh etc.).
Oylik foydalanish: 13 944, Oylik reyting: 4.0556 (36 ovoz)Medley Vox is an algorithm for separating multiple singers within a single music track and evaluation dataset for this task.
Vokal Oylik foydalanish: 7 886, Oylik reyting: 3.7857 (14 ovoz)MVSep Multichannel BS - uses the best vocal model to extract sound from multi-channel audio (5.1, 7.1, etc.).
Vokal Oylik foydalanish: 2 664, Oylik reyting: 4.0000 (4 ovoz)A model for separating male and female voices within a single vocal track. The track should contain only voices, no music.
Vokal Oylik foydalanish: 14 339, Oylik reyting: 3.5135 (37 ovoz)Choir Extraction Model
Vokal Oylik foydalanish: 2 727, Oylik reyting: 4.7500 (4 ovoz)Model to separate vocals and strings to SATB parts (Soprano, Alto, Tenor, and Bass)
Vokal Oylik foydalanish: 13 715, Oylik reyting: 3.2917 (24 ovoz)The MVSep Drums model produces high-quality separation of music into a drums part and everything else.
Barabanlar Oylik foydalanish: 20 673, Oylik reyting: 4.5510 (49 ovoz)The MVSep Bass model produces high-quality separation of music into a bass part and everything else.
Bas Oylik foydalanish: 13 036, Oylik reyting: 4.1795 (39 ovoz)Synth extraction model
Oylik foydalanish: 9 578, Oylik reyting: 4.8958 (48 ovoz)The DrumSep model divides the drum track into several types: 'kick', 'snare', 'toms', 'cymbals', 'hh', 'ride', 'crash' etc.
Barabanlar Oylik foydalanish: 24 622, Oylik reyting: 4.9170 (253 ovoz)MVSep Piano model is based on MDX23C, MelRoformer and SCNet Large architectures. It produces high quality separation for piano and other stems.
Pianino Klavishli asboblar Oylik foydalanish: 9 211, Oylik reyting: 4.0500 (20 ovoz)No data found
Pianino Klavishli asboblar Oylik foydalanish: 1 846, Oylik reyting: 5.0000 (55 ovoz)The MVSep Keys is a high quality model for separating music into keys instruments and everything else.
Klavishli asboblar Oylik foydalanish: 3 825, Oylik reyting: 4.6667 (21 ovoz)The MVSep Organ model produces high-quality separation of music into an organ part and everything else.
Klavishli asboblar Oylik foydalanish: 3 280, Oylik reyting: 4.3784 (37 ovoz)No data found
Klavishli asboblar Oylik foydalanish: 988, Oylik reyting: 4.9375 (16 ovoz)No data found
Klavishli asboblar Oylik foydalanish: 1 789, Oylik reyting: 4.3333 (9 ovoz)No data found
Klavishli asboblar Oylik foydalanish: 777, Oylik reyting: 4.4444 (9 ovoz)No data found
Klavishli asboblar Oylik foydalanish: 369, Oylik reyting: 4.6667 (3 ovoz)The Rhodes (Fender Rhodes) is an iconic electromechanical keyboard instrument designed by Harold Rhodes.
Klavishli asboblar Oylik foydalanish: 1 047, Oylik reyting: 4.7273 (11 ovoz)The Clavinet is an electromechanical keyboard instrument developed and manufactured by the German company Hohner from 1964 to 1982.
Klavishli asboblar Oylik foydalanish: 475, Oylik reyting: 4.5556 (9 ovoz)Mellotron is a electromechanical keyboard instrument developed in England in the early 1960s.
Klavishli asboblar Oylik foydalanish: 630, Oylik reyting: 4.5000 (8 ovoz)The MVSep Guitar model produces high-quality separation of music into a guitar part (including acoustic and electronic) and everything else.
Gitara Torli asboblar Oylik foydalanish: 9 225, Oylik reyting: 4.5263 (38 ovoz)No data found
Gitara Torli asboblar Oylik foydalanish: 5 999, Oylik reyting: 4.8636 (66 ovoz)No data found
Gitara Torli asboblar Oylik foydalanish: 6 152, Oylik reyting: 4.8765 (81 ovoz)No data found
Gitara Torli asboblar Oylik foydalanish: 15 596, Oylik reyting: 4.6636 (107 ovoz)Algorithm for isolating pedal steel guitar sounds from a melody
Gitara Oylik foydalanish: 825, Oylik reyting: 4.8462 (13 ovoz)The MVSep Plucked Strings is a high quality model for separating music into plucked strings instruments and everything else.
Torli asboblar Oylik foydalanish: 1 967, Oylik reyting: 4.8571 (7 ovoz)No data found
Torli asboblar Oylik foydalanish: 1 050, Oylik reyting: 5.0000 (3 ovoz)No data found
Torli asboblar Oylik foydalanish: 573, Oylik reyting: 5.0000 (1 ovoz)No data found
Torli asboblar Oylik foydalanish: 488, Oylik reyting: 5.0000 (3 ovoz)No data found
Torli asboblar Oylik foydalanish: 296, Oylik reyting: 5.0000 (1 ovoz)No data found
Torli asboblar Oylik foydalanish: 482, Oylik reyting: 4.7500 (4 ovoz)No data found
Torli asboblar Oylik foydalanish: 434, Oylik reyting: 4.6667 (3 ovoz)The MVSep Bowed Strings is a high quality model for separating music into bowed string instruments and everything else.
Torli asboblar Oylik foydalanish: 11 068, Oylik reyting: 4.9184 (98 ovoz)No data found
Torli asboblar Oylik foydalanish: 3 895, Oylik reyting: 4.7647 (17 ovoz)No data found
Torli asboblar Oylik foydalanish: 1 011, Oylik reyting: 3.4000 (5 ovoz)No data found
Torli asboblar Oylik foydalanish: 1 494, Oylik reyting: 4.3333 (6 ovoz)No data found
Torli asboblar Oylik foydalanish: 1 490, Oylik reyting: 3.6667 (3 ovoz)The MVSep Wind model produces high-quality separation of music into a wind part and everything else.
Nafasli asboblar Oylik foydalanish: 7 145, Oylik reyting: 4.8333 (24 ovoz)The MVSep Brass is a high quality model for separating music into brass wind instruments and everything else.
Nafasli asboblar Oylik foydalanish: 4 526, Oylik reyting: 4.7895 (19 ovoz)The MVSep Woodwind is a high quality model for separating music into woodwind instruments and everything else.
Nafasli asboblar Oylik foydalanish: 1 640, Oylik reyting: 4.6667 (3 ovoz)No data found
Nafasli asboblar Oylik foydalanish: 4 519, Oylik reyting: 4.4762 (21 ovoz)No data found
Nafasli asboblar Oylik foydalanish: 2 594, Oylik reyting: 5.0000 (11 ovoz)No data found
Nafasli asboblar Oylik foydalanish: 2 514, Oylik reyting: 4.5000 (18 ovoz)No data found
Nafasli asboblar Oylik foydalanish: 1 031, Oylik reyting: 4.2857 (7 ovoz)No data found
Nafasli asboblar Oylik foydalanish: 448, Oylik reyting: 0 (0 ovoz)No data found
Nafasli asboblar Oylik foydalanish: 932, Oylik reyting: 4.8333 (6 ovoz)No data found
Nafasli asboblar Oylik foydalanish: 789, Oylik reyting: 5.0000 (4 ovoz)No data found
Nafasli asboblar Oylik foydalanish: 632, Oylik reyting: 4.9286 (14 ovoz)No data found
Nafasli asboblar Oylik foydalanish: 540, Oylik reyting: 4.5000 (2 ovoz)No data found
Nafasli asboblar Oylik foydalanish: 390, Oylik reyting: 0 (0 ovoz)The MVSep Bagpipes model provides high-quality music separation, isolating the bagpipes from the rest of the track.
Nafasli asboblar Oylik foydalanish: 283, Oylik reyting: 4.7500 (4 ovoz)No data found
Nafasli asboblar Oylik foydalanish: 808, Oylik reyting: 5.0000 (2 ovoz)The MVSep Percussion is a high quality model for separating music into percussion instruments and everything else.
Perkussiya Oylik foydalanish: 5 110, Oylik reyting: 4.8148 (27 ovoz)No data found
Perkussiya Oylik foydalanish: 1 033, Oylik reyting: 5.0000 (11 ovoz)No data found
Perkussiya Oylik foydalanish: 497, Oylik reyting: 4.0000 (3 ovoz)No data found
Perkussiya Oylik foydalanish: 563, Oylik reyting: 4.5000 (4 ovoz)No data found
Perkussiya Oylik foydalanish: 627, Oylik reyting: 4.0000 (1 ovoz)No data found
Perkussiya Oylik foydalanish: 379, Oylik reyting: 5.0000 (4 ovoz)No data found
Perkussiya Oylik foydalanish: 1 067, Oylik reyting: 4.6667 (12 ovoz)No data found
Perkussiya Oylik foydalanish: 1 687, Oylik reyting: 5.0000 (10 ovoz)No data found
Perkussiya Oylik foydalanish: 312, Oylik reyting: 5.0000 (1 ovoz)No data found
Perkussiya Oylik foydalanish: 570, Oylik reyting: 4.6000 (5 ovoz)No data found
Perkussiya Oylik foydalanish: 418, Oylik reyting: 5.0000 (1 ovoz)No data found
Perkussiya Oylik foydalanish: 1 240, Oylik reyting: 3.4545 (11 ovoz)No data found
Perkussiya Oylik foydalanish: 661, Oylik reyting: 5.0000 (4 ovoz)No data found
Oylik foydalanish: 5 829, Oylik reyting: 1.5000 (2 ovoz)BandIt Plus model for separating tracks into speech, music and effects.
Oylik foydalanish: 2 123, Oylik reyting: 3.0000 (1 ovoz)Bandit v2 is a model for cinematic audio source separation in 3 stems: speech, music, effects/sfx. It was trained on DnR v3 dataset.
Oylik foydalanish: 3 063, Oylik reyting: 1.0000 (2 ovoz)MVSep DnR v3 is a cinematic model for splitting tracks into 3 stems: music, sfx and speech.
Oylik foydalanish: 27 603, Oylik reyting: 3.5000 (20 ovoz)The MVSep Braam model provides high-quality music separation, isolating the cinematic "Braam" sound effect from the rest of the track.
Oylik foydalanish: 347, Oylik reyting: 4.0000 (1 ovoz)Algorithm for isolating "Riser" sounds from a musical melody
Oylik foydalanish: 448, Oylik reyting: 0 (0 ovoz)No data found
Oylik foydalanish: 10 080, Oylik reyting: 4.5000 (26 ovoz)The algorithm restores the quality of audio. For example MP3 files compressed to 128 kbps or lower and other types.
Super rezolyutsiya Oylik foydalanish: 9 318, Oylik reyting: 2.9333 (15 ovoz)Set of different models to remove reverberation effect from music.
Oylik foydalanish: 17 711, Oylik reyting: 4.2941 (51 ovoz)No data found
Oylik foydalanish: 21 766, Oylik reyting: 4.8298 (141 ovoz)Algorithm AudioSR: Versatile Audio Super-resolution at Scale. Algorithm restores high frequencies.
Super rezolyutsiya Oylik foydalanish: 6 207, Oylik reyting: 4.2157 (51 ovoz)FlashSR - audio super resolution algorithm for restoring high frequencies
Super rezolyutsiya Oylik foydalanish: 2 169, Oylik reyting: 3.2857 (7 ovoz)Generating audio based on a given text prompt
Oylik foydalanish: 319, Oylik reyting: 5.0000 (26 ovoz)Whisper is a pre-trained model for automatic speech recognition (ASR) and speech translation.
Oylik foydalanish: 974, Oylik reyting: 3.0000 (2 ovoz)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.
Oylik foydalanish: 228, Oylik reyting: 0 (0 ovoz)open-source speech recognition (speech-to-text) model from the Qwen team
Oylik foydalanish: 299, Oylik reyting: 3.0000 (1 ovoz)VibeVoice - is a model for generating natural conversational dialogues from text with the ability to use a reference voice for cloning purposes.
Oylik foydalanish: 3 213, Oylik reyting: 2.3333 (9 ovoz)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.
Oylik foydalanish: 46, Oylik reyting: 0 (0 ovoz)Qwen3-TTS (Custom Voice) is a powerful speech generation model with predefined speakers
Oylik foydalanish: 42, Oylik reyting: 5.0000 (1 ovoz)Qwen3-TTS (Voice Design) is a powerful speech generation model capable of creating a voice based on a description.
Oylik foydalanish: 62, Oylik reyting: 5.0000 (2 ovoz)Qwen3-TTS (Voice Cloning) is a powerful speech generation model capable of cloning a voice based on a reference audio file
Oylik foydalanish: 371, Oylik reyting: 1.0000 (1 ovoz)A model for separating a musical composition into 53 different musical instruments
Vokal Barabanlar Bas Pianino Gitara Nafasli asboblar Torli asboblar Perkussiya Klavishli asboblar Oylik foydalanish: 10 812, Oylik reyting: 4.7692 (13 ovoz)Model to convert mono audio into true stereo
Oylik foydalanish: 781, Oylik reyting: 4.2222 (9 ovoz)Advanced AI Model for Stereo-to-Surround Upmixing
Oylik foydalanish: 2 926, Oylik reyting: 4.1556 (45 ovoz)A model that recovers pure dry lead vocals from processed audio
Oylik foydalanish: 9 199, Oylik reyting: 4.4872 (39 ovoz)A model for de-limiting songs. It properly restores the waveform and dynamics of the audio.
Oylik foydalanish: 649, Oylik reyting: 3.1667 (6 ovoz)Bark - is a generative model for realistic speech, music, and emotions.
Oylik foydalanish: 40, Oylik reyting: 1.0000 (1 ovoz)MVSep MultiSpeaker (MDX23C) - this model tries to isolate the most loud voice from all other voices.
Oylik foydalanish: 1 215, Oylik reyting: 1.3333 (6 ovoz)UniverSR - audio super resolution algorithm for restoring high frequencies
Super rezolyutsiya Oylik foydalanish: 2 068, Oylik reyting: 3.0000 (16 ovoz)Сlean vocals (acapellas) from artifacts after phase inversion
Oylik foydalanish: 2 082, Oylik reyting: 4.2857 (7 ovoz)The algorithm adds "whispering" effect to vocals.
Oylik foydalanish: 558, Oylik reyting: 4.0000 (1 ovoz)Algorithm for extracting the phantom center from a audio track
Oylik foydalanish: 2 261, Oylik reyting: 3.6667 (9 ovoz)Matchering is a novel tool for audio matching and mastering.
Oylik foydalanish: 3 322, Oylik reyting: 4.0000 (11 ovoz)SOME is a MIDI extractor that can convert singing voice to MIDI sequence.
MIDI Oylik foydalanish: 607, Oylik reyting: 0 (0 ovoz)High-quality transcription of piano music into MIDI
MIDI Oylik foydalanish: 2 773, Oylik reyting: 4.8000 (10 ovoz)Basic Pitch is a modern neural network that converts melodic audio recordings into notes (MIDI format).
MIDI Oylik foydalanish: 1 541, Oylik reyting: 3.7000 (10 ovoz)Automatic transcription of drum audio to MIDI
MIDI Oylik foydalanish: 989, Oylik reyting: 0 (0 ovoz)Multi-instrument MIDI transcription
MIDI Non-commercial Oylik foydalanish: 4 480, Oylik reyting: 4.2857 (14 ovoz)HeartMuLa - an open-source AI model for music generation (an alternative to Suno and Udio)
Oylik foydalanish: 148, Oylik reyting: 1.0000 (2 ovoz)Algorithm Demucs3 (A and B versions)
Vokal Barabanlar Bas Oylik foydalanish: 329, Oylik reyting: 0 (0 ovoz)No data found
Vokal Barabanlar Bas Oylik foydalanish: 127, Oylik reyting: 0 (0 ovoz)Experimental model VitLarge23 based on Vision Transformers. In terms of metrics, it is slightly inferior to the MDX23C, but may work better in some cases.
Vokal Oylik foydalanish: 96, Oylik reyting: 0 (0 ovoz)No data found
Vokal Oylik foydalanish: 146, Oylik reyting: 5.0000 (1 ovoz)No data found
Oylik foydalanish: 629, Oylik reyting: 1.5000 (2 ovoz)No data found
Vokal Oylik foydalanish: 67, Oylik reyting: 0 (0 ovoz)No data found
Vokal Barabanlar Bas Oylik foydalanish: 60, Oylik reyting: 0 (0 ovoz)No data found
Vokal Barabanlar Bas Oylik foydalanish: 48, Oylik reyting: 0 (0 ovoz)No data found
Vokal Barabanlar Bas Oylik foydalanish: 52, Oylik reyting: 0 (0 ovoz)No data found
Oylik foydalanish: 137, Oylik reyting: 0 (0 ovoz)No data found
Oylik foydalanish: 205, Oylik reyting: 0 (0 ovoz)No data found
Oylik foydalanish: 100, Oylik reyting: 0 (0 ovoz)The LarsNet model divides the drums stem into 5 types: 'kick', 'snare', 'cymbals', 'toms', 'hihat'.
Barabanlar Oylik foydalanish: 571, Oylik reyting: 1.0000 (1 ovoz) Ma’lumot topilmadiBugungi bepul ajratishlar: 50 / 50.
Navbatdagi qayta ishlanmagan fayllar: 300. Hozir GPU orqali qayta ishlanmoqda: 24
0:00
