Upload models.json with huggingface_hub

models.json

@@ -3,26 +3,376 @@
   "updated_at": "2026-04-10",
   "min_app_version": "1.0",
   "categories": [
+    {
+      "id": "segmentation",
+      "name": "Segmentation",
+      "icon": "person.and.background.dotted",
+      "order": 1
+    },
+    {
+      "id": "enhancement",
+      "name": "Image Enhancement",
+      "icon": "wand.and.stars",
+      "order": 2
+    },
+    {
+      "id": "detection",
+      "name": "Object Detection",
+      "icon": "viewfinder",
+      "order": 3
+    },
     {
       "id": "depth",
-      "name": "Monocular Depth Estimation",
+      "name": "Depth & Geometry",
       "icon": "cube.transparent",
-      "order": 1
+      "order": 4
+    },
+    {
+      "id": "vision_language",
+      "name": "Vision-Language",
+      "icon": "text.viewfinder",
+      "order": 5
+    },
+    {
+      "id": "face",
+      "name": "Face Processing",
+      "icon": "face.smiling",
+      "order": 6
+    },
+    {
+      "id": "generation",
+      "name": "Image Generation",
+      "icon": "sparkles",
+      "order": 7
+    },
+    {
+      "id": "video",
+      "name": "Video Processing",
+      "icon": "film",
+      "order": 8
+    },
+    {
+      "id": "audio",
+      "name": "Audio Processing",
+      "icon": "waveform.circle",
+      "order": 9
+    },
+    {
+      "id": "speech",
+      "name": "Speech & Music",
+      "icon": "music.note",
+      "order": 10
     }
   ],
   "models": [
+    {
+      "id": "rmbg_1_4",
+      "name": "RMBG-1.4",
+      "subtitle": "BRIA AI, 2023",
+      "category_id": "segmentation",
+      "description_md": "High-quality background removal. Outputs foreground with alpha mask. 1024×1024 input.",
+      "demo": {
+        "template": "image_in_out",
+        "config": {
+          "input_size": 1024,
+          "output_type": "mask"
+        }
+      },
+      "files": [
+        {
+          "name": "RMBG_1_4.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/rmbg/RMBG_1_4.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 38771210,
+          "sha256": "a80dbb5f04c922a8fa698c38592e4e52af4e62471d70bc7c59c28a3355a1da95",
+          "compute_units": "cpuAndGPU",
+          "kind": "model"
+        }
+      ],
+      "requirements": {
+        "min_ios": "17.0",
+        "min_ram_mb": 300
+      },
+      "license": {
+        "name": "Apache-2.0",
+        "url": "https://huggingface.co/briaai/RMBG-1.4"
+      },
+      "upstream": {
+        "name": "briaai/RMBG-1.4",
+        "url": "https://huggingface.co/briaai/RMBG-1.4",
+        "year": 2023
+      }
+    },
+    {
+      "id": "ddcolor",
+      "name": "DDColor Tiny",
+      "subtitle": "Image Colorization, 2023",
+      "category_id": "enhancement",
+      "description_md": "Automatic grayscale image colorization via dual decoders. 512×512 input.",
+      "demo": {
+        "template": "image_in_out",
+        "config": {
+          "input_size": 512,
+          "output_type": "image"
+        }
+      },
+      "files": [
+        {
+          "name": "DDColor_Tiny.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/ddcolor/DDColor_Tiny.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 212344570,
+          "sha256": "bfecea37d66005f602efe13978360b8e4707923234c3d1d00beeb4e36cb1b02c",
+          "compute_units": "all",
+          "kind": "model"
+        }
+      ],
+      "requirements": {
+        "min_ios": "17.0",
+        "min_ram_mb": 400
+      },
+      "license": {
+        "name": "Apache-2.0",
+        "url": "https://github.com/piddnad/DDColor"
+      },
+      "upstream": {
+        "name": "piddnad/DDColor",
+        "url": "https://github.com/piddnad/DDColor",
+        "year": 2023
+      }
+    },
+    {
+      "id": "sinsr",
+      "name": "SinSR",
+      "subtitle": "Single-Step Super-Resolution, 2024",
+      "category_id": "enhancement",
+      "description_md": "4× super-resolution via single-step diffusion. 256→1024. Swin Transformer denoiser (FP32).",
+      "demo": {
+        "template": "image_in_out",
+        "config": {
+          "input_size": 256,
+          "output_type": "image"
+        }
+      },
+      "files": [
+        {
+          "name": "SinSR_Encoder.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/sinsr/SinSR_Encoder.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 41246338,
+          "sha256": "fdec09d17561ec1bb5a2e829683d48c2b45e76b876285619a6e29a3523b8b7e2",
+          "compute_units": "cpuAndGPU",
+          "kind": "model"
+        },
+        {
+          "name": "SinSR_Denoiser.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/sinsr/SinSR_Denoiser.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 440014511,
+          "sha256": "b31374c2d539b2cdd81499d6062c801ca00e405f5a67507cd609d14e2d6d4beb",
+          "compute_units": "cpuOnly",
+          "kind": "model"
+        },
+        {
+          "name": "SinSR_Decoder.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/sinsr/SinSR_Decoder.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 60880285,
+          "sha256": "b8b9a7b52d6b240cf9fb3352b286ea83eb984fd73f5dd81c9f034f0016a5cb8c",
+          "compute_units": "cpuAndGPU",
+          "kind": "model"
+        }
+      ],
+      "requirements": {
+        "min_ios": "17.0",
+        "min_ram_mb": 600
+      },
+      "license": {
+        "name": "Apache-2.0",
+        "url": "https://github.com/wyf0912/SinSR"
+      },
+      "upstream": {
+        "name": "wyf0912/SinSR",
+        "url": "https://github.com/wyf0912/SinSR",
+        "year": 2024
+      }
+    },
+    {
+      "id": "yolo26s",
+      "name": "YOLO26s",
+      "subtitle": "NMS-Free Detection, 2026",
+      "category_id": "detection",
+      "description_md": "NMS-free object detection. 640×640 input, 80 COCO classes.",
+      "demo": {
+        "template": "image_detection",
+        "config": {
+          "input_size": 640,
+          "confidence_threshold": 0.25
+        }
+      },
+      "files": [
+        {
+          "name": "yolo26s.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/yolo26/yolo26s.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 17697581,
+          "sha256": "0ec02fb0cf2dbd6e09601cbbc00a9734156ea4c2a52b0da23a984337074c6fd4",
+          "compute_units": "all",
+          "kind": "model"
+        }
+      ],
+      "requirements": {
+        "min_ios": "17.0",
+        "min_ram_mb": 300
+      },
+      "license": {
+        "name": "AGPL-3.0",
+        "url": "https://github.com/ultralytics/ultralytics"
+      },
+      "upstream": {
+        "name": "ultralytics/ultralytics",
+        "url": "https://github.com/ultralytics/ultralytics",
+        "year": 2026
+      }
+    },
+    {
+      "id": "yolo11s",
+      "name": "YOLO11s",
+      "subtitle": "Object Detection, 2024",
+      "category_id": "detection",
+      "description_md": "YOLO11 small detection with Vision framework NMS. 640×640 input.",
+      "demo": {
+        "template": "image_detection",
+        "config": {
+          "input_size": 640,
+          "confidence_threshold": 0.25
+        }
+      },
+      "files": [
+        {
+          "name": "yolo11s.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/yolov9/yolo11s.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 17580204,
+          "sha256": "79e82aacc3ad20fc1eb990df6979fae9b927d4b06f33bd20ec0e1c0dcb7d1f6b",
+          "compute_units": "all",
+          "kind": "model"
+        }
+      ],
+      "requirements": {
+        "min_ios": "17.0",
+        "min_ram_mb": 300
+      },
+      "license": {
+        "name": "AGPL-3.0",
+        "url": "https://github.com/ultralytics/ultralytics"
+      },
+      "upstream": {
+        "name": "ultralytics/ultralytics",
+        "url": "https://github.com/ultralytics/ultralytics",
+        "year": 2024
+      }
+    },
+    {
+      "id": "yolov10n",
+      "name": "YOLOv10n",
+      "subtitle": "Object Detection, 2024",
+      "category_id": "detection",
+      "description_md": "YOLOv10 nano. 640×640 input. Dual-assignment strategy.",
+      "demo": {
+        "template": "image_detection",
+        "config": {
+          "input_size": 640,
+          "confidence_threshold": 0.25
+        }
+      },
+      "files": [
+        {
+          "name": "YOLOv10N.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/yolov10/YOLOv10N.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 4309168,
+          "sha256": "9a687144a6b0b764f508c8f544fe46b6674629b8f09a1e99d8ca69b0be899891",
+          "compute_units": "all",
+          "kind": "model"
+        }
+      ],
+      "requirements": {
+        "min_ios": "17.0",
+        "min_ram_mb": 300
+      },
+      "license": {
+        "name": "AGPL-3.0",
+        "url": "https://github.com/THU-MIG/yolov10"
+      },
+      "upstream": {
+        "name": "THU-MIG/yolov10",
+        "url": "https://github.com/THU-MIG/yolov10",
+        "year": 2024
+      }
+    },
+    {
+      "id": "yoloworld",
+      "name": "YOLO-World",
+      "subtitle": "Open-Vocabulary Detection, 2024",
+      "category_id": "detection",
+      "description_md": "Open-vocabulary detection. Type any text query. YOLO-World V2-S + CLIP ViT-B/32.",
+      "demo": {
+        "template": "open_vocab_detection",
+        "config": {
+          "input_size": 640
+        }
+      },
+      "files": [
+        {
+          "name": "yoloworld_detector.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/yoloworld/yoloworld_detector.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 23710620,
+          "sha256": "611d299ae74c83f90a5cc9f4585709859d5db735baa8ade721e0c2d99cd5af92",
+          "compute_units": "all",
+          "kind": "model"
+        },
+        {
+          "name": "clip_text_encoder.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/yoloworld/clip_text_encoder.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 116681932,
+          "sha256": "45770a743297e8c2a57cc330d4f5c80f47734263680895b33b593b50dd2c382b",
+          "compute_units": "cpuOnly",
+          "kind": "model"
+        }
+      ],
+      "requirements": {
+        "min_ios": "17.0",
+        "min_ram_mb": 600
+      },
+      "license": {
+        "name": "GPL-3.0",
+        "url": "https://github.com/AILab-CVC/YOLO-World"
+      },
+      "upstream": {
+        "name": "AILab-CVC/YOLO-World",
+        "url": "https://github.com/AILab-CVC/YOLO-World",
+        "year": 2024
+      }
+    },
     {
       "id": "moge2_vitb_normal_504",
       "name": "MoGe-2 ViT-B (504×504)",
       "subtitle": "Microsoft, CVPR 2025",
       "category_id": "depth",
-      "description_md": "Monocular geometry from a single image. Predicts metric depth, surface normals, and a confidence mask in one forward pass.\n\nBased on a DINOv2 ViT-B/14 backbone with three task heads. The successor to MiDaS-style relative depth: depth comes out in real meters.",
-      "thumbnail_url": null,
+      "description_md": "Monocular geometry from a single image. Metric depth, surface normals, confidence mask. DINOv2 ViT-B/14 backbone.",
       "demo": {
         "template": "depth_visualization",
         "config": {
           "input_size": 504,
-          "output_keys": ["depth", "normal", "mask", "metric_scale"],
+          "output_keys": [
+            "depth",
+            "normal",
+            "mask",
+            "metric_scale"
+          ],
           "depth_unit": "meters"
         }
       },
@@ -33,7 +383,8 @@
           "archive": "zip",
           "size_bytes": 193312088,
           "sha256": "f60cfb4804707a489d99e24453188cd31ddcabb299bbf6da4507edc9cecbf9e7",
-          "compute_units": "all"
+          "compute_units": "all",
+          "kind": "model"
         }
       ],
       "requirements": {
@@ -49,6 +400,652 @@
         "url": "https://github.com/microsoft/MoGe",
         "year": 2025
       }
+    },
+    {
+      "id": "siglip",
+      "name": "SigLIP",
+      "subtitle": "Zero-Shot Classification, 2023",
+      "category_id": "vision_language",
+      "description_md": "Zero-shot image classification. Dual encoder (image + text). 224×224 input.",
+      "demo": {
+        "template": "zero_shot_classify",
+        "config": {
+          "input_size": 224,
+          "image_encoder": "SigLIP_ImageEncoder.mlpackage.zip",
+          "text_encoder": "SigLIP_TextEncoder.mlpackage.zip",
+          "vocab_file": "siglip_vocab.json",
+          "prompt_template": "{}",
+          "logit_scale": 117.33
+        }
+      },
+      "files": [
+        {
+          "name": "SigLIP_ImageEncoder.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/siglip/SigLIP_ImageEncoder.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 170352400,
+          "sha256": "98f6abf5f4aa145199f4ae22305f9c1d5929eee6b126daad84783b2b2090ee24",
+          "compute_units": "cpuOnly",
+          "kind": "model"
+        },
+        {
+          "name": "SigLIP_TextEncoder.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/siglip/SigLIP_TextEncoder.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 203975769,
+          "sha256": "9dead2d58705838aef7ad83c3bf4036698c78d872ca1cdd04f2c4a6272009ccf",
+          "compute_units": "cpuOnly",
+          "kind": "model"
+        },
+        {
+          "name": "siglip_vocab.json",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/siglip/siglip_vocab.json",
+          "size_bytes": 673754,
+          "sha256": "b94b3a58e04f619936b3890804dff7c478522c07515ff748cf127c5443ee5229",
+          "kind": "vocab"
+        }
+      ],
+      "requirements": {
+        "min_ios": "17.0",
+        "min_ram_mb": 800
+      },
+      "license": {
+        "name": "Apache-2.0",
+        "url": "https://github.com/google-research/big_vision"
+      },
+      "upstream": {
+        "name": "google-research/big_vision",
+        "url": "https://github.com/google-research/big_vision",
+        "year": 2023
+      }
+    },
+    {
+      "id": "florence2",
+      "name": "Florence-2",
+      "subtitle": "Microsoft, 2024",
+      "category_id": "vision_language",
+      "description_md": "Vision-language captioning, OCR, and VQA. Three-stage encoder-decoder. 768×768 input.",
+      "demo": {
+        "template": "image_to_text",
+        "config": {
+          "image_size": 768,
+          "max_tokens": 256,
+          "vision_encoder": "Florence2VisionEncoder.mlpackage.zip",
+          "text_encoder": "Florence2TextEncoder.mlpackage.zip",
+          "decoder": "Florence2Decoder.mlpackage.zip",
+          "vocab_file": "florence2_vocab.json",
+          "tasks": {
+            "caption": [
+              0,
+              2264,
+              473,
+              5,
+              2274,
+              6190,
+              116,
+              2
+            ],
+            "detailed_caption": [
+              0,
+              2264,
+              473,
+              5,
+              31962,
+              2274,
+              6190,
+              116,
+              2
+            ],
+            "ocr": [
+              0,
+              2264,
+              473,
+              5,
+              71307,
+              116,
+              2
+            ]
+          }
+        }
+      },
+      "files": [
+        {
+          "name": "Florence2VisionEncoder.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/florence2/Florence2VisionEncoder.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 81198683,
+          "sha256": "9422f189c21220a0f9966eb9d780856772feb55597dcc579fc4e3c88990d0046",
+          "compute_units": "cpuOnly",
+          "kind": "model"
+        },
+        {
+          "name": "Florence2TextEncoder.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/florence2/Florence2TextEncoder.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 72742890,
+          "sha256": "f985deeef0408ea8aac33ac4f5c6d9635cd9c64c98b53f85031db6e27f3bfd92",
+          "compute_units": "cpuOnly",
+          "kind": "model"
+        },
+        {
+          "name": "Florence2Decoder.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/florence2/Florence2Decoder.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 85329746,
+          "sha256": "fe85a6faab5281272bcd79dabfbf87d60ba1a78dd9455e2bf71c67a134d61dc5",
+          "compute_units": "cpuOnly",
+          "kind": "model"
+        },
+        {
+          "name": "florence2_vocab.json",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/florence2/florence2_vocab.json",
+          "size_bytes": 999352,
+          "sha256": "861fee9af5520403f6dbb4940d6af6627f1481b71cdc4a870f1f61344e57e645",
+          "kind": "vocab"
+        }
+      ],
+      "requirements": {
+        "min_ios": "17.0",
+        "min_ram_mb": 1200
+      },
+      "license": {
+        "name": "MIT",
+        "url": "https://huggingface.co/microsoft/Florence-2-base"
+      },
+      "upstream": {
+        "name": "microsoft/Florence-2",
+        "url": "https://huggingface.co/microsoft/Florence-2-base",
+        "year": 2024
+      }
+    },
+    {
+      "id": "face3d",
+      "name": "3DDFA V2",
+      "subtitle": "3D Face Reconstruction, 2020",
+      "category_id": "face",
+      "description_md": "Single-image 3D face reconstruction. Predicts 6 DoF pose + expression parameters.",
+      "demo": {
+        "template": "face_3d",
+        "config": {
+          "input_size": 120
+        }
+      },
+      "files": [
+        {
+          "name": "3DDFA_V2.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/face3d/3DDFA_V2.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 6083375,
+          "sha256": "0f715dc220c046f558e3b8fc65246df9a2eec77182830a16628783430cdacdc8",
+          "compute_units": "all",
+          "kind": "model"
+        }
+      ],
+      "requirements": {
+        "min_ios": "17.0",
+        "min_ram_mb": 200
+      },
+      "license": {
+        "name": "MIT",
+        "url": "https://github.com/cleardusk/3DDFA_V2"
+      },
+      "upstream": {
+        "name": "cleardusk/3DDFA_V2",
+        "url": "https://github.com/cleardusk/3DDFA_V2",
+        "year": 2020
+      }
+    },
+    {
+      "id": "hypersd",
+      "name": "Hyper-SD (1-Step)",
+      "subtitle": "ByteDance, 2024",
+      "category_id": "generation",
+      "description_md": "Single-step text-to-image from SD1.5 via TCD distillation. 512×512. Chunked UNet (6-bit).",
+      "demo": {
+        "template": "text_to_image",
+        "config": {
+          "image_size": 512,
+          "latent_size": 64,
+          "latent_channels": 4,
+          "steps": 1,
+          "guidance_scale": 1.0,
+          "text_encoder": "HyperSDTextEncoder.mlpackage.zip",
+          "unet_chunk1": "HyperSDUnetChunk1.mlpackage.zip",
+          "unet_chunk2": "HyperSDUnetChunk2.mlpackage.zip",
+          "vae_decoder": "HyperSDVAEDecoder.mlpackage.zip",
+          "vocab_file": "vocab.json",
+          "merges_file": "merges.txt"
+        }
+      },
+      "files": [
+        {
+          "name": "HyperSDTextEncoder.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/hypersd/HyperSDTextEncoder.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 226397794,
+          "sha256": "201b0fcc3573811aac6a4e8545c695bc4fb2f7710ea0d60c227919d87b37687e",
+          "compute_units": "cpuAndNeuralEngine",
+          "kind": "model"
+        },
+        {
+          "name": "HyperSDUnetChunk1.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/hypersd/HyperSDUnetChunk1.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 324819653,
+          "sha256": "279da11b8231aeeb9045f6ceabebb3a68c20a1b86ecc81aa6914b77ce76d5203",
+          "compute_units": "cpuAndNeuralEngine",
+          "kind": "model"
+        },
+        {
+          "name": "HyperSDUnetChunk2.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/hypersd/HyperSDUnetChunk2.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 304530429,
+          "sha256": "0a700d11a105da589bb3e5666e38b9c72fa283149951b253fc11722e70e72faa",
+          "compute_units": "cpuAndNeuralEngine",
+          "kind": "model"
+        },
+        {
+          "name": "HyperSDVAEDecoder.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/hypersd/HyperSDVAEDecoder.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 91282754,
+          "sha256": "1260371542d845a2261ed2de36c5fe3e9ccb740a6ceb59b1990705d125e8cf66",
+          "compute_units": "cpuAndNeuralEngine",
+          "kind": "model"
+        },
+        {
+          "name": "vocab.json",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/hypersd/vocab.json",
+          "size_bytes": 1059962,
+          "sha256": "e089ad92ba36837a0d31433e555c8f45fe601ab5c221d4f607ded32d9f7a4349",
+          "kind": "vocab"
+        },
+        {
+          "name": "merges.txt",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/hypersd/merges.txt",
+          "size_bytes": 524619,
+          "sha256": "9fd691f7c8039210e0fced15865466c65820d09b63988b0174bfe25de299051a",
+          "kind": "vocab"
+        }
+      ],
+      "requirements": {
+        "min_ios": "17.0",
+        "min_ram_mb": 1000
+      },
+      "license": {
+        "name": "OpenRAIL-M",
+        "url": "https://huggingface.co/ByteDance/Hyper-SD"
+      },
+      "upstream": {
+        "name": "ByteDance/Hyper-SD",
+        "url": "https://huggingface.co/ByteDance/Hyper-SD",
+        "year": 2024
+      }
+    },
+    {
+      "id": "matanyone",
+      "name": "MatAnyone",
+      "subtitle": "Video Matting, 2025",
+      "category_id": "video",
+      "description_md": "Temporally consistent video matting. 5-model pipeline with memory propagation.",
+      "demo": {
+        "template": "video_matting",
+        "config": {
+          "frame_size": 512,
+          "encoder": "MatAnyone_encoder.mlpackage.zip",
+          "mask_encoder": "MatAnyone_mask_encoder.mlpackage.zip",
+          "read_first": "MatAnyone_read_first.mlpackage.zip",
+          "read": "MatAnyone_read.mlpackage.zip",
+          "decoder": "MatAnyone_decoder.mlpackage.zip"
+        }
+      },
+      "files": [
+        {
+          "name": "MatAnyone_encoder.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/matanyone/MatAnyone_encoder.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 17306121,
+          "sha256": "97ffd6bc4611f9a3351dc890fc00954ba48171e517e66a39f7a5f1f38110dfda",
+          "compute_units": "all",
+          "kind": "model"
+        },
+        {
+          "name": "MatAnyone_mask_encoder.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/matanyone/MatAnyone_mask_encoder.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 16819866,
+          "sha256": "ba67559188ffc64d8e46418c051c6a55815d4482def17519fa518daac7d5a911",
+          "compute_units": "all",
+          "kind": "model"
+        },
+        {
+          "name": "MatAnyone_read_first.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/matanyone/MatAnyone_read_first.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 21991849,
+          "sha256": "34daf7227dbcec7373a3fef175259fa7ec631ed8cb91d5595ca57ee9b22df7bb",
+          "compute_units": "cpuOnly",
+          "kind": "model"
+        },
+        {
+          "name": "MatAnyone_read.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/matanyone/MatAnyone_read.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 22135429,
+          "sha256": "052e52c0ffb7ff9ede448128950cd4c1c9a96589b6900c82b5104d99addb7fa5",
+          "compute_units": "cpuOnly",
+          "kind": "model"
+        },
+        {
+          "name": "MatAnyone_decoder.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/matanyone/MatAnyone_decoder.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 8807630,
+          "sha256": "67136aa67000e604838fe9aa7de151c514ef84f0b83f1da0f043cf70652d28eb",
+          "compute_units": "all",
+          "kind": "model"
+        }
+      ],
+      "requirements": {
+        "min_ios": "17.0",
+        "min_ram_mb": 800
+      },
+      "license": {
+        "name": "MIT",
+        "url": "https://github.com/pq-yang/MatAnyone"
+      },
+      "upstream": {
+        "name": "pq-yang/MatAnyone",
+        "url": "https://github.com/pq-yang/MatAnyone",
+        "year": 2025
+      }
+    },
+    {
+      "id": "demucs",
+      "name": "HTDemucs",
+      "subtitle": "Audio Source Separation",
+      "category_id": "audio",
+      "description_md": "Split music into 4 stems: drums, bass, vocals, other. 44.1 kHz stereo, FP32.",
+      "demo": {
+        "template": "audio_in_out",
+        "config": {
+          "sample_rate": 44100,
+          "segment_length": 343980,
+          "output_stems": [
+            "drums",
+            "bass",
+            "vocals",
+            "other"
+          ]
+        }
+      },
+      "files": [
+        {
+          "name": "HTDemucs_SourceSeparation_F32.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/demucs/HTDemucs_SourceSeparation_F32.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 79076395,
+          "sha256": "0fbb941e15a5b2fa425d14fe630ed4c14b6dee72780c1f5b2b05f58803bce5f7",
+          "compute_units": "cpuOnly",
+          "kind": "model"
+        }
+      ],
+      "requirements": {
+        "min_ios": "17.0",
+        "min_ram_mb": 1000
+      },
+      "license": {
+        "name": "MIT",
+        "url": "https://github.com/adefossez/demucs"
+      },
+      "upstream": {
+        "name": "adefossez/demucs",
+        "url": "https://github.com/adefossez/demucs",
+        "year": 2021
+      }
+    },
+    {
+      "id": "kokoro",
+      "name": "Kokoro-82M",
+      "subtitle": "Multilingual TTS",
+      "category_id": "speech",
+      "description_md": "English + Japanese text-to-speech. 24 kHz. StyleTTS2 + iSTFTNet vocoder. Multiple voices.",
+      "demo": {
+        "template": "text_to_audio",
+        "config": {
+          "mode": "tts",
+          "sample_rate": 24000,
+          "vocab_file": "kokoro_vocab.json",
+          "voices": [
+            "af_heart",
+            "af_bella",
+            "am_michael",
+            "bf_emma",
+            "bm_george"
+          ]
+        }
+      },
+      "files": [
+        {
+          "name": "Kokoro_Predictor.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/kokoro/Kokoro_Predictor.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 72191470,
+          "sha256": "af1d55dc842980c32b5591a70f603941f11ab60a435bed0c13a107a8ef467bed",
+          "compute_units": "cpuAndGPU",
+          "kind": "model"
+        },
+        {
+          "name": "Kokoro_Decoder_128.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/kokoro/Kokoro_Decoder_128.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 229120589,
+          "sha256": "cece0d072f5ba6aa3f729cf4c76b4de51823bcc65a26ab363c10441c3cd8b306",
+          "compute_units": "all",
+          "kind": "model"
+        },
+        {
+          "name": "Kokoro_Decoder_256.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/kokoro/Kokoro_Decoder_256.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 229123438,
+          "sha256": "36d5e16d5c5ccb500fc96f1b07a1d5ac57b791f8e09e61b78319d76949003efe",
+          "compute_units": "all",
+          "kind": "model"
+        },
+        {
+          "name": "Kokoro_Decoder_512.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/kokoro/Kokoro_Decoder_512.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 229128735,
+          "sha256": "0a44484c327e4fe8443b0bcf104d6964fe3f30d628c9e78aee3f31af7f2475dc",
+          "compute_units": "all",
+          "kind": "model"
+        },
+        {
+          "name": "kokoro_vocab.json",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/kokoro/kokoro_vocab.json",
+          "size_bytes": 1144,
+          "sha256": "70abefbe8a1c8865e43e0a43bbdc25b91a33e4aa053479d443ccf23e20a59e5d",
+          "kind": "vocab"
+        }
+      ],
+      "requirements": {
+        "min_ios": "17.0",
+        "min_ram_mb": 1000
+      },
+      "license": {
+        "name": "Apache-2.0",
+        "url": "https://huggingface.co/hexgrad/Kokoro-82M"
+      },
+      "upstream": {
+        "name": "hexgrad/Kokoro-82M",
+        "url": "https://huggingface.co/hexgrad/Kokoro-82M",
+        "year": 2024
+      }
+    },
+    {
+      "id": "stable_audio",
+      "name": "Stable Audio Open",
+      "subtitle": "Text-to-Music, 2024",
+      "category_id": "speech",
+      "description_md": "Text-to-music. Up to 11.9s stereo 44.1 kHz. Rectified flow DiT + T5 + Oobleck VAE.",
+      "demo": {
+        "template": "text_to_audio",
+        "config": {
+          "mode": "music",
+          "sample_rate": 44100,
+          "max_duration": 11.9
+        }
+      },
+      "files": [
+        {
+          "name": "StableAudioT5Encoder.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/stableaudio/StableAudioT5Encoder.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 98538259,
+          "sha256": "319a8ba775d309240253ced68a03a3923d0aec9a79f608044f9403bdcfe4b741",
+          "compute_units": "cpuOnly",
+          "kind": "model"
+        },
+        {
+          "name": "StableAudioNumberEmbedder.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/stableaudio/StableAudioNumberEmbedder.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 376018,
+          "sha256": "04bdc5de00a2cf1c4a18f80c94f0d74ecfab41f3ad99f2fb7a031d6ff5af75da",
+          "compute_units": "cpuOnly",
+          "kind": "model"
+        },
+        {
+          "name": "StableAudioDiT.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/stableaudio/StableAudioDiT.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 1265748504,
+          "sha256": "b17da4fc4df857821d39dbdf7d3bfe7062a2272ab3e5df1284d545afb54047e4",
+          "compute_units": "cpuOnly",
+          "kind": "model"
+        },
+        {
+          "name": "StableAudioVAEDecoder.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/stableaudio/StableAudioVAEDecoder.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 144960275,
+          "sha256": "7207544cca9799cc1d6803c5e81badd0bb4527b2d3a64d5cab5700a5f19a9374",
+          "compute_units": "cpuAndGPU",
+          "kind": "model"
+        },
+        {
+          "name": "t5_vocab.json",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/stableaudio/t5_vocab.json",
+          "size_bytes": 749757,
+          "sha256": "7c9ff3ac1b3dbcaa617ee659f2df68688cfd44f1a5eb3be3fa0a2f56c749d56a",
+          "kind": "vocab"
+        }
+      ],
+      "requirements": {
+        "min_ios": "17.0",
+        "min_ram_mb": 1200
+      },
+      "license": {
+        "name": "custom",
+        "url": "https://huggingface.co/stabilityai/stable-audio-open-small"
+      },
+      "upstream": {
+        "name": "stabilityai/stable-audio-open-small",
+        "url": "https://huggingface.co/stabilityai/stable-audio-open-small",
+        "year": 2024
+      }
+    },
+    {
+      "id": "openvoice",
+      "name": "OpenVoice V2",
+      "subtitle": "Voice Cloning",
+      "category_id": "audio",
+      "description_md": "Zero-shot voice conversion. Clone a speaker from ~10s reference audio.",
+      "demo": {
+        "template": "audio_in_out",
+        "config": {
+          "sample_rate": 22050,
+          "output_stems": [
+            "converted"
+          ]
+        }
+      },
+      "files": [
+        {
+          "name": "OpenVoice_SpeakerEncoder.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/openvoice/OpenVoice_SpeakerEncoder.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 1519880,
+          "sha256": "c3f2a96aaf5ecb5c5afc62b3d3dfbd47dc7ae64bc3edb7aa68befb54aef74459",
+          "compute_units": "cpuAndGPU",
+          "kind": "model"
+        },
+        {
+          "name": "OpenVoice_VoiceConverter.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/openvoice/OpenVoice_VoiceConverter.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 59799630,
+          "sha256": "ef3ce8a2d1564aefa13830d7d0ca43f85e0aa62d5f59622c8bc456c307ab5e05",
+          "compute_units": "cpuAndGPU",
+          "kind": "model"
+        }
+      ],
+      "requirements": {
+        "min_ios": "17.0",
+        "min_ram_mb": 500
+      },
+      "license": {
+        "name": "MIT",
+        "url": "https://github.com/myshell-ai/OpenVoice"
+      },
+      "upstream": {
+        "name": "myshell-ai/OpenVoice",
+        "url": "https://github.com/myshell-ai/OpenVoice",
+        "year": 2023
+      }
+    },
+    {
+      "id": "diarization",
+      "name": "Pyannote Diarization",
+      "subtitle": "Speaker Identification",
+      "category_id": "audio",
+      "description_md": "Speaker diarization: who spoke when. 16 kHz mono, 10s segments.",
+      "demo": {
+        "template": "audio_in_out",
+        "config": {
+          "sample_rate": 16000,
+          "output_stems": [
+            "speaker_timeline"
+          ]
+        }
+      },
+      "files": [
+        {
+          "name": "SpeakerSegmentation.mlpackage.zip",
+          "url": "https://huggingface.co/mlboydaisuke/coreml-zoo/resolve/main/diarization/SpeakerSegmentation.mlpackage.zip",
+          "archive": "zip",
+          "size_bytes": 5327137,
+          "sha256": "dcfa2b98900f2b99029abfb593644b70418186a6ec2e94c9a79c2b3d7a84378a",
+          "compute_units": "cpuAndGPU",
+          "kind": "model"
+        }
+      ],
+      "requirements": {
+        "min_ios": "17.0",
+        "min_ram_mb": 200
+      },
+      "license": {
+        "name": "MIT",
+        "url": "https://github.com/pyannote/pyannote-audio"
+      },
+      "upstream": {
+        "name": "pyannote/pyannote-audio",
+        "url": "https://github.com/pyannote/pyannote-audio",
+        "year": 2021
+      }
     }
   ]
-}
+}