Leaderboard
The most-downloaded open-weight models on Hugging Face over the last 30 days — the community's working set, straight from the Hub and refreshed every 12 hours.
| # | Model | Downloads | Likes | Task |
|---|---|---|---|---|
| 1 | sentence-transformers/all-MiniLM-L6-v2 | 183.8M | 6K | sentence similarity |
| 2 | cross-encoder/ms-marco-MiniLM-L6-v2 | 67.7M | 437 | text ranking |
| 3 | BAAI/bge-small-en-v1.5 BAAI | 50.5M | 662 | feature extraction |
| 4 | BAAI/bge-m3 BAAI | 30.3M | 4K | sentence similarity |
| 5 | Comfy-Org/MiniMax-H3 | 30.2M | 2.5K | — |
| 6 | google-bert/bert-base-uncased | 29.2M | 3.1K | fill mask |
| 7 | openai-community/gpt2 | 29M | 3.9K | text generation |
| 8 | sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2 | 28.9M | 1.5K | sentence similarity |
| 9 | argmaxinc/whisperkit-coreml | 24M | 260 | automatic speech recognition |
| 10 | google-t5/t5-small | 23.7M | 687 | translation |
| 11 | Qwen/Qwen3-0.6B Qwen · Alibaba | 23.4M | 1.9K | text generation |
| 12 | autogluon/chronos-2 | 22.7M | 94 | time series forecasting |
| 13 | sentence-transformers/all-mpnet-base-v2 | 20.1M | 1.4K | sentence similarity |
| 14 | google/electra-base-discriminator Google DeepMind | 19.6M | 218 | — |
| 15 | openai/clip-vit-base-patch32 | 17.6M | 1.3K | zero shot image classification |
| 16 | jonatasgrosman/wav2vec2-large-xlsr-53-japanese | 16.3M | 118 | automatic speech recognition |
| 17 | google/vit-base-patch16-224 Google DeepMind | 15.7M | 1K | image classification |
| 18 | timm/mobilenetv3_small_100.lamb_in1k | 15.1M | 156 | image classification |
| 19 | BAAI/bge-reranker-v2-m3 BAAI | 14.6M | 1.3K | text classification |
| 20 | Qwen/Qwen3-14B Qwen · Alibaba | 13M | 529 | text generation |
| 21 | nomic-ai/nomic-embed-text-v1.5 | 11.9M | 988 | sentence similarity |
| 22 | hexgrad/Kokoro-82M | 11.2M | 7.3K | text to speech |
| 23 | Comfy-Org/Qwen-Image-2.1 | 10.6M | 1.2K | text to image |
| 24 | prithivMLmods/Gender-Classifier-Mini | 10.5M | 19 | image classification |
| 25 | amazon/chronos-2 | 10.3M | 562 | time series forecasting |
Leaderboard data is cached from Hugging Face Hub and refreshed every 12 hours. Scores shown are for reference; see the source for full methodology.