Most used papers

  1. Qwen3 Technical Report 2399 models, 117,116,472 downloads in 30 days, 152 upvotes
  2. Soaring from 4K to 400K: Extending LLM's Context with Activation Beacon 138 models, 87,681,795 downloads in 30 days, 30 upvotes
  3. YaRN: Efficient Context Window Extension of Large Language Models 1730 models, 80,216,972 downloads in 30 days, 88 upvotes
  4. Qwen2 Technical Report 1580 models, 53,896,306 downloads in 30 days, 142 upvotes
  5. Chronos: Learning the Language of Time Series 83 models, 48,502,107 downloads in 30 days, 33 upvotes
  6. Gemma 4 Technical Report 281 models, 37,421,562 downloads in 30 days, 63 upvotes
  7. Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution 610 models, 35,514,616 downloads in 30 days, 63 upvotes
  8. OpenMed NER: Open-Source, Domain-Adapted State-of-the-Art Transformers for Biomedical NER Across 12 Public Datasets 1249 models, 34,914,139 downloads in 30 days, 7 upvotes
  9. Chronos-2: From Univariate to Universal Forecasting 21 models, 30,208,790 downloads in 30 days, 14 upvotes
  10. Multilingual E5 Text Embeddings: A Technical Report 1036 models, 26,275,783 downloads in 30 days, 23 upvotes
  11. Qwen2.5-VL Technical Report 409 models, 22,905,440 downloads in 30 days, 146 upvotes
  12. Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Models 356 models, 17,552,566 downloads in 30 days, 55 upvotes
  13. Nomic Embed: Training a Reproducible Long Context Text Embedder 41 models, 15,808,249 downloads in 30 days, 19 upvotes
  14. GLM-5: from Vibe Coding to Agentic Engineering 389 models, 12,038,585 downloads in 30 days, 94 upvotes
  15. SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features 847 models, 11,835,895 downloads in 30 days, 118 upvotes
  16. Qwen2.5-Coder Technical Report 434 models, 11,099,189 downloads in 30 days, 111 upvotes
  17. GPQA: A Graduate-Level Google-Proof Q&A Benchmark 603 models, 8,412,062 downloads in 30 days, 38 upvotes
  18. Gemini: A Family of Highly Capable Multimodal Models 624 models, 8,083,444 downloads in 30 days, 50 upvotes
  19. BLINK: Multimodal Large Language Models Can See but Not Perceive 471 models, 7,808,924 downloads in 30 days, 20 upvotes
  20. Qwen3-TTS Technical Report 369 models, 7,658,974 downloads in 30 days, 54 upvotes
  21. Nemotron 3 Nano: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning 114 models, 6,189,194 downloads in 30 days, 28 upvotes
  22. NVIDIA Nemotron 3: Efficient and Open Intelligence 103 models, 6,176,224 downloads in 30 days, 27 upvotes
  23. Smarter, Better, Faster, Longer: A Modern Bidirectional Encoder for Fast, Memory Efficient, and Long Context Finetuning and Inference 112 models, 5,286,555 downloads in 30 days, 105 upvotes
  24. DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning 517 models, 4,749,565 downloads in 30 days, 271 upvotes
  25. LTX-2: Efficient Joint Audio-Visual Foundation Model 188 models, 4,449,090 downloads in 30 days, 110 upvotes
  26. SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis 163 models, 4,419,693 downloads in 30 days, 94 upvotes
  27. SmolLM2: When Smol Goes Big -- Data-Centric Training of a Small Language Model 112 models, 4,320,313 downloads in 30 days, 158 upvotes
  28. Florence-2: Advancing a Unified Representation for a Variety of Vision Tasks 88 models, 3,864,942 downloads in 30 days, 98 upvotes
  29. EmbeddingGemma: Powerful and Lightweight Text Representations 297 models, 3,697,590 downloads in 30 days, 33 upvotes
  30. MiniCPM4: Ultra-Efficient LLMs on End Devices 100 models, 3,581,930 downloads in 30 days, 78 upvotes
  31. mGTE: Generalized Long-Context Text Representation and Reranking Models for Multilingual Text Retrieval 39 models, 3,550,113 downloads in 30 days, 16 upvotes
  32. DINOv3 114 models, 3,351,196 downloads in 30 days, 194 upvotes
  33. Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data 36 models, 3,253,851 downloads in 30 days, 64 upvotes
  34. Improving Text Embeddings with Large Language Models 35 models, 3,187,121 downloads in 30 days, 84 upvotes
  35. Qwen3-VL-Embedding and Qwen3-VL-Reranker: A Unified Framework for State-of-the-Art Multimodal Retrieval and Ranking 26 models, 3,048,115 downloads in 30 days, 45 upvotes
  36. GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Models 152 models, 3,008,593 downloads in 30 days, 144 upvotes
  37. Qwen3-ASR Technical Report 110 models, 3,004,915 downloads in 30 days, 33 upvotes
  38. Depth Anything V2 59 models, 2,926,548 downloads in 30 days, 83 upvotes
  39. How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites 322 models, 2,832,557 downloads in 30 days, 47 upvotes
  40. InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks 321 models, 2,826,467 downloads in 30 days, 22 upvotes
  41. Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling 295 models, 2,825,433 downloads in 30 days, 103 upvotes
  42. DeepSeek-OCR: Contexts Optical Compression 52 models, 2,817,888 downloads in 30 days, 63 upvotes
  43. Data Science and Technology Towards AGI Part I: Tiered Data Management 80 models, 2,565,404 downloads in 30 days, 5 upvotes
  44. Voxtral Realtime 30 models, 2,339,042 downloads in 30 days, 15 upvotes
  45. Structured 3D Latents for Scalable and Versatile 3D Generation 56 models, 2,253,599 downloads in 30 days, 36 upvotes
  46. Simple and Controllable Music Generation 106 models, 2,235,846 downloads in 30 days, 170 upvotes
  47. Scaling Open-Vocabulary Object Detection 24 models, 2,178,060 downloads in 30 days, 16 upvotes
  48. Qwen2.5-1M Technical Report 371 models, 2,100,380 downloads in 30 days, 51 upvotes
  49. jina-embeddings-v3: Multilingual Embeddings With Task LoRA 14 models, 2,097,177 downloads in 30 days, 19 upvotes
  50. Qwen-Image Technical Report 167 models, 2,091,958 downloads in 30 days, 173 upvotes
  51. RULER: What's the Real Context Size of Your Long-Context Language Models? 334 models, 2,074,639 downloads in 30 days, 28 upvotes
  52. SmolVLM: Redefining small and efficient multimodal models 35 models, 1,992,380 downloads in 30 days, 158 upvotes
  53. ChatQA: Building GPT-4 Level Conversational QA Models 43 models, 1,890,389 downloads in 30 days, 35 upvotes
  54. YuE: Scaling Open Foundation Models for Long-Form Music Generation 28 models, 1,833,502 downloads in 30 days, 57 upvotes
  55. Llama 2: Open Foundation and Fine-Tuned Chat Models 566 models, 1,833,400 downloads in 30 days, 252 upvotes
  56. Perception Encoder: The best visual embeddings are not at the output of the network 112 models, 1,817,542 downloads in 30 days, 31 upvotes
  57. Z-Image: An Efficient Image Generation Foundation Model with Single-Stream Diffusion Transformer 113 models, 1,755,537 downloads in 30 days, 163 upvotes
  58. MOSEL: 950,000 Hours of Speech Data for Open-Source Speech Foundation Model Training on EU Languages 21 models, 1,726,910 downloads in 30 days, 14 upvotes
  59. Nemotron 3 Nano Omni: Efficient and Open Multimodal Intelligence 23 models, 1,669,562 downloads in 30 days, 17 upvotes
  60. Nemotron Elastic: Towards Efficient Many-in-One Reasoning LLMs 25 models, 1,608,799 downloads in 30 days, 22 upvotes
  61. Qwen Technical Report 299 models, 1,570,965 downloads in 30 days, 39 upvotes
  62. Decoupled DMD: CFG Augmentation as the Spear, Distribution Matching as the Shield 87 models, 1,565,647 downloads in 30 days, 22 upvotes
  63. DFlash: Block Diffusion for Flash Speculative Decoding 239 models, 1,565,244 downloads in 30 days, 41 upvotes
  64. MiniCPM-V: A GPT-4V Level MLLM on Your Phone 70 models, 1,553,994 downloads in 30 days, 69 upvotes
  65. Multimodal RewardBench 2: Evaluating Omni Reward Models for Interleaved Text and Image 2 models, 1,536,544 downloads in 30 days, 12 upvotes
  66. GLiNER2: An Efficient Multi-Task Information Extraction System with Schema-Driven Interface 28 models, 1,476,313 downloads in 30 days, 13 upvotes
  67. Training-Free Long-Context Scaling of Large Language Models 309 models, 1,470,561 downloads in 30 days, 23 upvotes
  68. EXAONE 3.5: Series of Large Language Models for Real-world Use Cases 20 models, 1,461,665 downloads in 30 days, 45 upvotes
  69. MInference 1.0: Accelerating Pre-filling for Long-Context LLMs via Dynamic Sparse Attention 298 models, 1,436,154 downloads in 30 days, 22 upvotes
  70. Enhancing the Reasoning Ability of Multimodal Large Language Models via Mixed Preference Optimization 226 models, 1,428,202 downloads in 30 days, 61 upvotes
  71. InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models 213 models, 1,339,722 downloads in 30 days, 239 upvotes
  72. Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone 124 models, 1,335,578 downloads in 30 days, 222 upvotes
  73. IndexCache: Accelerating Sparse Attention via Cross-Layer Index Reuse 161 models, 1,331,752 downloads in 30 days, 51 upvotes
  74. Unlimited OCR Works 47 models, 1,295,572 downloads in 30 days, 43 upvotes
  75. Gemma 3 Technical Report 724 models, 1,245,980 downloads in 30 days, 40 upvotes
  76. AfriMed-QA: A Pan-African, Multi-Specialty, Medical Question-Answering Benchmark Dataset 59 models, 1,198,487 downloads in 30 days, 3 upvotes
  77. MedXpertQA: Benchmarking Expert-Level Medical Reasoning and Understanding 59 models, 1,198,487 downloads in 30 days, 19 upvotes
  78. Vision Transformers Need Registers 96 models, 1,151,559 downloads in 30 days, 86 upvotes
  79. Olmo 3 44 models, 1,141,901 downloads in 30 days, 22 upvotes
  80. DSpark: Confidence-Scheduled Speculative Decoding with Semi-Autoregressive Generation 24 models, 1,122,425 downloads in 30 days, 33 upvotes
  81. Optimize Weight Rounding via Signed Gradient Descent for the Quantization of LLMs 437 models, 1,075,048 downloads in 30 days, 16 upvotes
  82. LFM2 Technical Report 243 models, 1,023,294 downloads in 30 days, 34 upvotes
  83. VibeVoice Technical Report 80 models, 988,752 downloads in 30 days, 118 upvotes
  84. Multimodal Latent Language Modeling with Next-Token Diffusion 74 models, 987,738 downloads in 30 days, 38 upvotes
  85. PaliGemma 2: A Family of Versatile VLMs for Transfer 76 models, 980,414 downloads in 30 days, 109 upvotes
  86. MedGemma Technical Report 51 models, 974,545 downloads in 30 days, 14 upvotes
  87. Wan: Open and Advanced Large-Scale Video Generative Models 240 models, 973,289 downloads in 30 days, 44 upvotes
  88. s1: Simple test-time scaling 89 models, 967,018 downloads in 30 days, 97 upvotes
  89. Improved Distribution Matching Distillation for Fast Image Synthesis 27 models, 960,752 downloads in 30 days, 10 upvotes
  90. DeepSeekMoE: Towards Ultimate Expert Specialization in Mixture-of-Experts Language Models 64 models, 960,419 downloads in 30 days, 65 upvotes
  91. HunyuanOCR-1.5: Making Lightweight OCR VLMs Faster and Better 3 models, 937,716 downloads in 30 days, 7 upvotes
  92. jina-reranker-v3: Last but Not Late Interaction for Document Reranking 5 models, 931,779 downloads in 30 days, 5 upvotes
  93. TranslateGemma Technical Report 27 models, 924,291 downloads in 30 days, 19 upvotes
  94. jina-embeddings-v5-omni: Text-Geometry-Preserving Multimodal Embeddings via Frozen-Tower Composition 28 models, 887,789 downloads in 30 days, 10 upvotes
  95. Kimi K2.5: Visual Agentic Intelligence 101 models, 873,868 downloads in 30 days, 219 upvotes
  96. InternVL3.5: Advancing Open-Source Multimodal Models in Versatility, Reasoning, and Efficiency 146 models, 866,428 downloads in 30 days, 170 upvotes
  97. Hermes 3 Technical Report 70 models, 858,740 downloads in 30 days, 24 upvotes
  98. SAM 2: Segment Anything in Images and Videos 77 models, 835,584 downloads in 30 days, 89 upvotes
  99. Ministral 3 81 models, 831,384 downloads in 30 days, 44 upvotes
  100. JaColBERTv2.5: Optimising Multi-Vector Retrievers to Create State-of-the-Art Japanese Retrievers with Constrained Resources 11 models, 817,701 downloads in 30 days, 21 upvotes

Data: hysts-bot-data/daily-papers-stats and the Daily Papers API. Open data: tardellirs/paper-pulse-data. Sister project: Model Pulse, the download history of every model on the Hub.