Catalog
Open-source model catalog
56 open-weight models with license, context window, features and the GPU memory they need at INT4. Specifications are synced from Hugging Face; open a model for the full sheet.
| Model | Params | Type | License | Context | Features | Languages | VRAM (INT4) | Downloads |
|---|---|---|---|---|---|---|---|---|
| Qwen3-8B Qwenauto-imported | 8.2B | Dense | apache-2.0 | 41K | toolsreasoning1 quants | EN | ~5 GB | 12,931,851 |
| Qwen2.5-7B-Instruct Qwenauto-imported | 7.6B | Dense | apache-2.0 | 33K | tools5 quants | EN | ~5 GB | 9,726,568 |
| Qwen3.6-35B-A3B-NVFP4 nvidiaauto-imported | 18.7B 0.6B active | MoE | apache-2.0 | 262K | toolsvisionreasoning3 quants | EN | ~12 GB | 8,277,959 |
| Qwen2.5-1.5B-Instruct Qwenauto-imported | 1.5B | Dense | apache-2.0 | 33K | tools5 quants | EN | ~1 GB | 7,239,606 |
| Qwen3-4B Qwenauto-imported | 4B | Dense | apache-2.0 | 41K | toolsreasoning2 quants | EN | ~3 GB | 7,098,338 |
| OTel-2.0-LLM-31B-IT farbodtavakkoliauto-imported | 31.3B | Dense | apache-2.0 | 262K | vision2 quants | EN | ~21 GB | 6,916,354 |
| gpt-oss-20b openaiauto-imported | 20.9B 2.6B active | MoE | apache-2.0 | 131K | 2 quants | EN | ~14 GB | 6,708,152 |
| gpt-oss-120b openaiauto-imported | 116.8B 3.7B active | MoE | apache-2.0 | 131K | 2 quants | EN | ~77 GB | 5,117,998 |
| Qwen2.5-3B-Instruct Qwenauto-imported | 3.1B | Dense | other non-commercial | 33K | tools5 quants | EN | ~2 GB | 5,090,920 |
| Qwen3-32B Qwenauto-imported | 32.8B | Dense | apache-2.0 | 41K | toolsreasoning1 quants | EN | ~22 GB | 5,035,207 |
| dolphin-2.9.1-yi-1.5-34b dphnauto-imported | 34.4B | Dense | apache-2.0 | 8K | tools4 quants | EN | ~23 GB | 4,820,992 |
| DeepSeek-V4-Flash-0731 deepseek-aiauto-imported | 304.2B 7.1B active | MoE | mit | 1049K | 3 quants | EN | ~201 GB | 4,178,593 |
| Qwen-72B Qwenauto-imported | 72.3B | Dense | other non-commercial | 33K | 2 quants | ZH EN | ~48 GB | 4,169,684 |
| Qwen3-4B-Instruct-2507 Qwenauto-imported | 4B | Dense | apache-2.0 | 262K | tools4 quants | EN | ~3 GB | 4,024,479 |
| Qwen3-1.7B Qwenauto-imported | 2B | Dense | apache-2.0 | 41K | toolsreasoning3 quants | EN | ~1 GB | 3,782,599 |
| Kimi-K3-DSpark RadixArkauto-imported | 2.2B | Dense | unknown non-commercial | 1049K | 2 quants | EN | ~1 GB | 3,550,124 |
| NVIDIA-Nemotron-3-Nano-4B-BF16 nvidiaauto-imported | 4B | Dense | other non-commercial | 262K | toolsreasoning2 quants | EN | ~3 GB | 3,485,601 |
| Qwen2.5-Coder-7B-Instruct Qwenauto-imported | 7.6B | Dense | apache-2.0 | 33K | tools6 quants | EN | ~5 GB | 2,629,675 |
| DeepSeek-V3.2 deepseek-aiauto-imported | 685.4B 21.4B active | MoE | mit | 164K | 3 quants | EN | ~452 GB | 2,421,883 |
| Qwen2.5-14B-Instruct Qwenauto-imported | 14.8B | Dense | apache-2.0 | 33K | tools6 quants | EN | ~10 GB | 2,352,910 |
| Qwen2.5-32B-Instruct Alibaba | 32.8B | Dense | Apache-2.0 | 128K | tools6 quants | EN ZH UR AR FR DE +2 | ~22 GB | 2,234,705 |
| Qwen3-14B Qwenauto-imported | 14.8B | Dense | apache-2.0 | 41K | toolsreasoning3 quants | EN | ~10 GB | 1,863,931 |
| GLM-4.7-Flash zai-orgauto-imported | 31.2B 2B active | MoE | mit | 203K | 4 quants | EN ZH | ~21 GB | 1,861,863 |
| Qwen2.5-Coder-14B-Instruct Qwenauto-imported | 14.8B | Dense | apache-2.0 | 33K | tools5 quants | EN | ~10 GB | 1,836,347 |
| Gemma-4-26B-A4B-NVFP4 nvidiaauto-imported | 14.4B | MoE | apache-2.0 | 262K | vision1 quants | EN | ~10 GB | 1,807,367 |
| Qwen3-30B-A3B Qwenauto-imported | 30.5B 1.9B active | MoE | apache-2.0 | 41K | toolsreasoning3 quants | EN | ~20 GB | 1,793,417 |
| Mistral-7B-Instruct-v0.2 mistralaiauto-imported | 7.2B | Dense | apache-2.0 | 33K | 4 quants | EN | ~5 GB | 1,782,269 |
| Gemma-4-31B-IT-NVFP4 nvidiaauto-imported | 20.9B | Dense | other non-commercial | 262K | vision2 quants | EN | ~14 GB | 1,638,603 |
| DeepSeek-V4-Flash deepseek-aiauto-imported | 290.9B 6.8B active | MoE | mit | 1049K | 2 quants | EN | ~192 GB | 1,590,858 |
| Qwen3-4B-Base Qwenauto-imported | 4B | Dense | apache-2.0 | 33K | toolsreasoning4 quants | EN | ~3 GB | 1,541,245 |
| Qwen3-1.7B-Base Qwenauto-imported | 1.7B | Dense | apache-2.0 | 33K | toolsreasoning4 quants | EN | ~1 GB | 1,535,756 |
| MiniMax-M2.7 MiniMaxAIauto-imported | 228.7B 7.1B active | MoE | other non-commercial | 205K | 2 quants | EN | ~151 GB | 1,461,696 |
| TinyLlama-1.1B-Chat-v1.0 TinyLlamaauto-imported | 1.1B | Dense | apache-2.0 | 2K | 5 quants | EN | ~1 GB | 1,460,358 |
| Qwen3.5-122B-A10B-NVFP4 nvidiaauto-imported | 64.6B 2B active | MoE | apache-2.0 | 262K | toolsvisionreasoning2 quants | EN | ~43 GB | 1,428,051 |
| Qwen3-Coder-Next-FP8 Qwenauto-imported | 79.7B 1.6B active | MoE | apache-2.0 | 262K | toolsreasoning1 quants | EN | ~53 GB | 1,423,926 |
| Qwen2.5-Coder-32B-Instruct Qwenauto-imported | 32.8B | Dense | apache-2.0 | 33K | tools6 quants | EN | ~22 GB | 1,348,917 |
| PowerMoE-3b ibm-researchauto-imported | 3.4B 0.7B active | MoE | apache-2.0 | 4K | 3 quants | EN | ~2 GB | 1,283,351 |
| NVIDIA-Nemotron-3-Super-120B-A12B-BF16 nvidiaauto-imported | 123.6B 5.3B active | MoE | other non-commercial | 262K | 2 quants | EN FR ES IT DE JA +1 | ~82 GB | 1,275,676 |
| Qwen3.8-27B-OBLITERATED OBLITERATUSauto-imported | 27.8B | Dense | apache-2.0 | 262K | visionreasoning4 quants | EN | ~18 GB | 1,256,591 |
| Ornith-1.5-35B-A3B-NVFP4 ornith-aiauto-imported | 19.5B 0.6B active | MoE | mit | 262K | toolsvisionreasoning1 quants | EN | ~13 GB | 1,167,934 |
| DeepSeek-V3 deepseek-aiauto-imported | 684.5B 21.4B active | MoE | unknown non-commercial | 164K | tools | EN | ~452 GB | 1,131,639 |
| Qwen3-Coder-30B-A3B-Instruct-FP8 Qwenauto-imported | 30.5B 1.9B active | MoE | apache-2.0 | 262K | toolsreasoning | EN | ~20 GB | 1,126,082 |
| Qwen2.5-VL-32B-Instruct Alibaba | 33.5B | Dense | Apache-2.0 | 128K | vision5 quants | EN ZH AR FR DE ES | ~22 GB | 1,121,845 |
| NVIDIA-Nemotron-3.5-Lightning-30B-A3B-NVFP4 nvidiaauto-imported | 17.8B 0.8B active | MoE | other non-commercial | 1049K | EN ES FR DE IT JA | ~12 GB | 1,084,597 | |
| Bonsai-27B-mlx-1bit prism-mlauto-imported | 1.7B | Dense | apache-2.0 | 262K | vision | EN | ~1 GB | 1,064,336 |
| Ternary-Bonsai-27B-mlx-2bit prism-mlauto-imported | 27.4B | Dense | apache-2.0 | 262K | vision | EN | ~18 GB | 1,056,895 |
| Llama-3.1-8B-Instruct-4bit mlx-communityauto-imported | 8B | Dense | llama3.1 | 131K | tools | EN DE FR IT PT HI +2 | ~5 GB | 1,046,370 |
| DeepSeek-V4-Flash-DSpark deepseek-aiauto-imported | 165.3B 3.9B active | MoE | mit | 1049K | EN | ~109 GB | 1,038,083 | |
| DeepSeek-V3-0324 deepseek-aiauto-imported | 684.5B 21.4B active | MoE | mit | 164K | tools | EN | ~452 GB | 991,371 |
| JiRackUltra_14b CMSManhattanauto-imported | 14.8B | Dense | mit | 131K | EN ZH JA KO FR ES +7 | ~10 GB | 991,317 | |
| GLM-5.2-FP8 zai-orgauto-imported | 753.3B 23.5B active | MoE | mit | 1049K | EN ZH | ~497 GB | 986,105 | |
| GLM-5.2 zai-orgauto-imported | 753.3B 23.5B active | MoE | mit | 1049K | EN ZH | ~497 GB | 974,216 | |
| GLM-5.3 zai-orgauto-imported | 753.3B 23.5B active | MoE | other non-commercial | 1049K | EN ZH | ~497 GB | 947,261 | |
| Qwen3-4B-Instruct-2507-FP8 Qwenauto-imported | 4.4B | Dense | apache-2.0 | 262K | tools | EN | ~3 GB | 940,305 |
| Llama 3.3 70B Instruct Meta | 70.6B | Dense | Llama 3.3 Community | 128K | 5 quants | EN FR DE HI IT PT +2 | ~47 GB | 916,029 |
| Mistral Small 3 (24B) Mistral AI | 23.6B | Dense | Apache-2.0 | 33K | 5 quants | EN FR DE ES IT PT +4 | ~16 GB | 48,741 |