Skip to content

All on-device models

Audio classification models for iPhone and Mac (Core ML & MLX)

27 models for this task are tagged Core ML or MLX on Hugging Face, or come from Apple: 22 Core ML, 5 MLX and 0 other Apple models. The table lists the most downloaded in the last 30 days.

Last updated 2026-11-02 (UTC).

Showing 27 of 27.

ModelSizeDownloads (30 days)
aufklarer/WeSpeaker-ResNet34-LM-MLXMLXfrom pyannote/wespeaker-voxceleb-resnet34-LM26.5 MB12.4K
aufklarer/Silero-VAD-v6.2.1-MLXMLX1.2 MB10.7K
aufklarer/Qwen3-ForcedAligner-0.6B-4bitMLXfrom Qwen/Qwen3-ForcedAligner-0.6B983 MB7,183
desert-ant-labs/uhmCore ML—5,609
aufklarer/Pyannote-Community-1-CoreMLCore MLfrom pyannote/speaker-diarization-community-1—1,543
mlx-community/silero-vadMLX—1,456
FluidInference/campplus-coremlCore ML—1,060
mlx-community/silero-vad-v6MLX—875
aufklarer/WeSpeaker-ResNet34-LM-CoreMLCore MLfrom pyannote/wespeaker-voxceleb-resnet34-LM—431
aufklarer/Smart-Turn-v3.2-CoreMLCore MLfrom pipecat-ai/smart-turn-v3—370
aufklarer/CamPlusPlus-Speaker-CoreMLCore ML—307
aufklarer/ReDimNet2-B6-CoreMLCore ML—278
aufklarer/Sortformer-Diarization-CoreMLCore MLfrom nvidia/diar_streaming_sortformer_4spk-v2.1—253
aufklarer/Qwen3-ForcedAligner-0.6B-CoreML-INT8Core MLfrom Qwen/Qwen3-ForcedAligner-0.6B—165
FluidInference/ls-eend-coremlCore ML—162
aufklarer/Qwen3-ForcedAligner-0.6B-CoreML-FP16Core MLfrom Qwen/Qwen3-ForcedAligner-0.6B—113
donpark/tyger-modelsCore MLfrom FluidInference/parakeet-redux-coreml—109
theoracleguy/Pyannote-Community-1-CoreMLCore MLfrom pyannote/speaker-diarization-community-1—104
GradientDescent2718/ls-eend-coremlCore ML—88
aufklarer/SpeechBrain-ECAPA-VoxLingua107-21M-CoreMLCore MLfrom speechbrain/lang-id-voxlingua107-ecapa—85
emee-ai/Sortformer-DiarizationCore MLfrom nvidia/diar_streaming_sortformer_4spk-v2.1—50
emee-ai/Qwen3-ForcedAligner-0.6B-INT8Core MLfrom Qwen/Qwen3-ForcedAligner-0.6B—49
aufklarer/Qwen3-ForcedAligner-0.6B-CoreML-INT4Core MLfrom Qwen/Qwen3-ForcedAligner-0.6B—45
aufklarer/KWS-Zipformer-3M-CoreML-INT8Core ML—40
Yehor/YAMNet-CoreMLCore ML—32
bwarzecha/axii-diarization-coremlCore ML—20
NagaYu/bleep-0.09bCore ML—17

Reading the table

  • Downloads are Hugging Face's count for the last 30 days, not all time. Hugging Face counts requests for a model's main file, so one install can count more than once.
  • Size is the total of the repo's current files on Hugging Face, in the sizes it shows (1 GB = 1,000 MB). Some repos hold several variants of a model, so it can be larger than the one download you need; a dash means it has not been measured yet.
  • Trending is Hugging Face's own score for recent attention. "from" under a name is the model it was made from.
  • Parameters is the parameter count in the model files. A count marked * comes from the model's name (such as 27B) instead: the files gave none, or a much smaller one, as packed low-bit files can.
  • Quantization is read from the model's tags and name, not checked against the files.

Before you use one

The list comes from Hugging Face's public API and is updated daily. We don't test each model. Check the license and read the model card before you ship one in an app: some licenses limit commercial use, and "other" means the terms are on the model card. Models whose names or tags point at adult or safety-removed content are left out, and so are gated models that need a Hugging Face sign-in to download.

Other tasks