| 1 | Underdog-Bark-3.0-CleanConwayResearch | Speech Recognition | 44 | — | 66 | — |
| 2 | Spark-X2.6-1.7BXHToken | Text Generation | 43 | — | 45 | 1.7B |
| 3 | Raon-1-mini-preview-ggufKRAFTON | Text Generation | 43 | — | 66 | — |
| 4 | MemT-1.0-FlashMemTensor | Text Generation | 43 | — | 114 | 288B |
| 5 | Qwen3.8-27B-OBLITERATEDOBLITERATUS | Text Generation | 43 | ▲14.4%(up) | 1,577 | 27.8B |
| 6 | Krea-2-Rawkrea | Text to Image | 42 | ▲5.2%(up) | 1,194 | 12.8B |
| 7 | Minimax-h3-Turbolightx2v | — | 42 | ▲22.4%(up) | 1,163 | — |
| 8 | gemma-4-31B-itgoogle | Image Text to Text | 42 | ▲4.0%(up) | 4,225 | 31.3B |
| 9 | Kahya-TTS-v1.0AlicanKiraz0 | Text to Speech | 42 | — | 42 | 2.3B |
| 10 | LFM2.5-24B-A2BLiquidAI | Text Generation | 42 | — | 587 | 24B |
| 11 | FLUX.1-schnellblack-forest-labs | Text to Image | 41 | ▲0.4%(up) | 6,487 | 11.9B |
| 12 | MiniMax-M3.1-Flash-GGUFunsloth | Image Text to Text | 41 | — | 41 | — |
| 13 | d3-ultravllm-sr | Zero-Shot Classification | 41 | — | 41 | 26.1B |
| 14 | FLUX.2-klein-base-9b-fp8black-forest-labs | Image To Image | 41 | ▲4.0%(up) | 1,499 | — |
| 15 | Cyber-Tiel-Coder-35B-A3B-GGUF-MTPpeculiar-ragdoll | Image Text to Text | 41 | ▲145.4%(up) | 299 | — |
| 16 | Qwen-Image-2.1-Uncensored-GGUFKasugaiSakura | Text to Image | 41 | — | 46 | — |
| 17 | Qwen-Image-2.1Comfy-Org | — | 41 | — | 1,216 | — |
| 18 | Krea-2-Turbokrea | Text to Image | 41 | ▲5.5%(up) | 1,800 | 12.8B |
| 19 | Hy-SimulMT-30B-A3Btencent | Translation | 40 | — | 40 | 30.4B |
| 20 | keys-DeepSeek-V4.1-Flash-1017-Abliterateddrowzeys | Image Text to Text | 40 | — | 115 | — |
| 21 | granite-speech-5.0-2bibm-granite | Speech Recognition | 40 | — | 61 | 2.3B |
| 22 | Phonon-2.5FermionResearch | Speech Recognition | 40 | — | 40 | — |
| 23 | FLUX.1-devblack-forest-labs | Text to Image | 40 | ▲0.4%(up) | 15.8K | 11.9B |
| 24 | Qwen3.8-Flash-Next-NVFP4nvidia | Image Text to Text | 40 | ▲72.3%(up) | 422 | 120B |
| 25 | penclaw-GLM-5.3-abliteratedaudnai | Text Generation | 40 | — | 496 | 753B |
| 26 | Huihui-DeepSeek-V4.1-Flash-1017-abliterated-GGUFhuihui-ai | Image Text to Text | 39 | — | 50 | — |
| 27 | Groot-19B-A3B-261023GrootTeam | Image Text to Text | 39 | — | 40 | 19.4B |
| 28 | Swift-1.5-Qwen3.8-27B-GSQ-RCO-GGUFukisai | Text Generation | 39 | — | 450 | — |
| 29 | segmentation-3.0pyannote | Voice Activity Detection | 39 | ▲0.3%(up) | 2,421 | — |
| 30 | MOSS-Image-CharRef7err4 | — | 38 | — | 38 | — |
| 31 | Sage-A1.5-Q4_K_M-GGUFInternScience | Text Generation | 38 | — | 66 | — |
| 32 | Nanbeige4.3-33B-UNCENSORED-GGUFdealignai | Text Generation | 38 | — | 39 | — |
| 33 | Qwen3.8-27B-Uncensored-GGUForcarouter | Image Text to Text | 38 | ▲9.3%(up) | 1,625 | — |
| 34 | phonon-40m-engradium | Text to Speech | 38 | — | 90 | 50.5M |
| 35 | NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4nvidia | Image Text to Text | 38 | — | 244 | 65.8B |
| 36 | Qwen3.5-4BQwen | Image Text to Text | 38 | ▲3.5%(up) | 1,154 | 4.7B |
| 37 | Qwen3.5-9BQwen | Image Text to Text | 37 | ▲4.6%(up) | 2,267 | 9.7B |
| 38 | MiniMax-H3-PTD-LoRAstaobao-content-ai | Image To Video | 37 | — | 50 | — |
| 39 | rh-qwen-image-2.1v2.0-loraRunningHubAI | Text to Image | 36 | — | 36 | — |
| 40 | LFM3-VL-2B-GGUFLiquidAI | Image Text to Text | 36 | — | 36 | — |
| 41 | Huihui-Qwen3.8-27B-abliteratedhuihui-ai | Image Text to Text | 36 | ▲10.5%(up) | 652 | 27.8B |
| 42 | phonon-40m-multilingualgradium | Text to Speech | 36 | — | 72 | 50.5M |
| 43 | Nanbeige4.3-33B-GSQ-RCO-GGUFabenzerps | Text Generation | 35 | — | 35 | — |
| 44 | Kokoro-82Mhexgrad | Text to Speech | 35 | ▲2.4%(up) | 7,296 | — |
| 45 | ThinkingCap-Qwen3.8-Flash-Nextbottlecapai | Image Text to Text | 35 | — | 35 | 180B |
| 46 | Qwen3.8-Flash-Next-GGUFunsloth | Image Text to Text | 35 | ▲29.9%(up) | 1,280 | — |
| 47 | personaplex-7b-v1nvidia | Audio to Audio | 34 | ▲1.8%(up) | 2,913 | 8.4B |
| 48 | hobbylm-1B-computer-useHobbylm | Text Generation | 34 | — | 34 | — |
| 49 | DeepSeek-V4.1-Flash-1017-GGUFunsloth | Image Text to Text | 34 | — | 154 | — |
| 50 | Qwen3.8-Flash-Next-abliterated-GSQ-RCO-Strata-GGUFalesha-pro | Image Text to Text | 33 | — | 202 | — |