Measured by open model pull

Open models

How often an open AI model is downloaded.

#NameTrenddownloads in 30 daysGrowth

1tencent/Hy3text generation model13.2k → 192k14×

2CMSManhattan/JiRackDeltaNet_27btext generation model78k → 553k7.1×

3poolside/Laguna-XS.2text generation model27k → 188k7.0×

4allenai/Olmo-3-7B-Thinktext generation model52.8k → 364k6.9×

5poolside/Laguna-XS-2.1text generation model18.4k → 119k6.5×

6zai-org/GLM-5text generation model52.4k → 307k5.8×

7coder3101/Qwen3-VL-32B-Instruct-heretic-v2image text to text model10.7k → 58.8k5.5×

8ibm-granite/granite-guardian-4.1-8btext generation model10.3k → 54.2k5.3×

9poolside/Laguna-S-2.1text generation model27.5k → 108k3.9×

10CMSManhattan/JiRackUltra_1btext generation model647k → 2.5M3.9×

11Comfy-Org/Ming-Imagecomfy.org20.1k → 77.6k3.9×

12Viggle/Qwen-Image-2.1-viggle-turbotext to image model60.4k → 218k3.6×

13replicate/flash-attn3replicate.com31.2k → 112k3.6×

14zai-org/GLM-5.1text generation model73.9k → 229k3.1×

15datalab-to/chandraimage text to text model133k → 343k2.6×

16nvidia/Nemotron-3-Diarizationvoice activity detection model16k → 40.9k2.6×

17AiAngelGallery/ComfyPod-Models91.3k → 224k2.4×

18tue-mps/eomt-dinov3-coco-panoptic-base-640image segmentation model19.8k → 48.1k2.4×

19zenosai/MonkeyOCRv2-B-Parsingimage text to text model57.8k → 139k2.4×

20jinaai/jina-embeddings-v3-hffeature extraction model46.2k → 108k2.3×

21Datadog/Toto-2.0-313mtime series forecasting model38.3k → 89.1k2.3×

22zai-org/GLM-5.3-Flash-BF16image text to text model26.1k → 56.7k2.2×

23opendatalab/MinerU2.5-Pro-2605-1.2Bimage text to text model68.7k → 148k2.1×

24wfzyx/vonzero shot classification model36.9k → 78.1k2.1×

25XiaomiMiMo/MiMo-V2.6-Flash-RLtext generation model22.2k → 44.7k2.0×

At least as big as the median of its peers · 7 days · Downloads, not users. Models released in the last year; copies and converted versions are left out.