1,226 models · 1,226 open weights · 200 API-only · refreshed nightly
Open-weights models
Models whose weights are published on Hugging Face — download them and run them on your own hardware. Many are also served by hosted providers; those carry an "Also via API" badge with live prices.
| # | Model | Params | Context | Commercial use | 30d | Min VRAM |
|---|---|---|---|---|---|---|
| 51 | DeepSeek-V3 | 684.5 B | 164 K | unknown | 1.3 M | from 860.6 GB |
| 52 | NVIDIA-Nemotron-3-Nano-30B-A3B-NVFP4 | 18.2 B | 262 K | other | 1.3 M | from 24.5 GB |
| 53 | Qwen3-4B-Instruct-2507-FP8 | 4.4 B | 262 K | ✓ apache-2.0 | 1.3 M | from 6.9 GB |
| 54 | NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 | 31.6 B | 262 K | other | 1.3 M | from 74.7 GB |
| 55 | DeepSeek-R1-Distill-Qwen-1.5B | 1.8 B | 131 K | ✓ mit | 1.3 M | from 4.7 GB |
| 56 | Qwen2.5-32B-Instruct | 32.8 B | 33 K | ✓ apache-2.0 | 1.3 M | from 77.5 GB |
| 57 | MiniCPM5-2B | 2.5 B | 131 K | ✓ apache-2.0 | 1.3 M | from 6.4 GB |
| 58 | Qwen2.5-0.5B | 490 M | 33 K | ✓ apache-2.0 | 1.2 M | from 1.7 GB |
| 59 | Qwen3.8-35B-A3B-Distill-GGUF | 35.0 B | — | ✓ apache-2.0 | 1.2 M | from 19.6 GB |
| 60 | Ornith-1.5-397B-GGUF | — | — | ✓ mit | 1.2 M | from 1.5 GB |
| 61 | Qwen3-Coder-30B-A3B-Instruct-FP8 | 30.5 B | 262 K | ✓ apache-2.0 | 1.1 M | from 39.4 GB |
| 62 | Ornith-1.5-35B-A3B-NVFP4 | 19.5 B | — | ✓ mit | 1.1 M | from 29.2 GB |
| 63 | deepseek-v4-gguf | — | — | ✓ mit | 1.1 M | from 4.7 GB |
| 64 | SmolLM2-135M-Instruct | 130 M | 8 K | ✓ apache-2.0 | 1.1 M | from 0.8 GB |
| 65 | Llama-3.2-3B-Instruct | 3.2 B | — | ⚠ llama3.2 | 1.1 M | from 8.0 GB |
| 66 | MiniCPM5-2B-DSpark | 324 M | 131 K | ✓ apache-2.0 | 1.1 M | from 1.3 GB |
| 67 | Qwen3-30B-A3B-Instruct-2507 | 30.5 B | 262 K | ✓ apache-2.0 | 1.1 M | from 72.3 GB |
| 68 | NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 | 560.5 B | 262 K | other | 1.1 M | from 1,317.7 GB |
| 69 | Gemma-4-26B-A4B-NVFP4 | 14.4 B | — | ✓ apache-2.0 | 1.1 M | from 23.3 GB |
| 70 | GLM-5 | 753.9 B | 203 K | ✓ mit | 1.0 M | from 1,772.1 GB |
| 71 | gemma-3-1b-it | 1.0 B | — | ⚠ gemma | 1.0 M | from 2.8 GB |
| 72 | Qwen3-30B-A3B | 30.5 B | 41 K | ✓ apache-2.0 | 1.0 M | from 72.3 GB |
| 73 | Qwen3.8-27B-Heretic-Abliterated-Uncensored-GGUF | 27.0 B | — | ✓ apache-2.0 | 1.0 M | from 12.9 GB |
| 74 | bloomz-560m | 560 M | — | ⚠ bigscience-bloom-rail-1.0 | 997 K | from 1.8 GB |
| 75 | NVIDIA-Nemotron-3-Super-120B-A12B-BF16 | 123.6 B | 262 K | other | 978 K | from 291.0 GB |
| 76 | Qwen2.5-Coder-32B-Instruct-AWQ | 32.8 B | 33 K | ✓ apache-2.0 | 973 K | from 26.7 GB |
| 77 | GLM-4.7-Flash | 31.2 B | 203 K | ✓ mit | 972 K | from 73.9 GB |
| 78 | Ternary-Bonsai-27B-mlx-2bit | 2.6 B | — | ✓ apache-2.0 | 967 K | from 10.2 GB |
| 79 | TinyLlama-1.1B-Chat-v1.0 | 1.1 B | 2 K | ✓ apache-2.0 | 964 K | from 3.1 GB |
| 80 | Bonsai-27B-mlx-1bit | 1.7 B | — | ✓ apache-2.0 | 957 K | from 6.4 GB |
| 81 | Qwen2.5-32B-Instruct-GGUF | 32.0 B | 33 K | ✓ apache-2.0 | 955 K | from 15.2 GB |
| 82 | Qwen3-Coder-Next-FP8 | 79.7 B | 262 K | ✓ apache-2.0 | 950 K | from 100.9 GB |
| 83 | Ornith-1.0-35B-FP8 | 35.1 B | — | ✓ mit | 945 K | from 47.2 GB |
| 84 | NVIDIA-Nemotron-3.5-Lightning-30B-A3B-FP8 | 31.6 B | 262 K | unknown | 928 K | from 40.8 GB |
| 85 | NVIDIA-Nemotron-3.5-Lightning-30B-A3B-NVFP4 | 17.8 B | 1.0 M | other | 922 K | from 26.9 GB |
| 86 | Qwen2.5-32B-Instruct-AWQ | 32.8 B | 33 K | ✓ apache-2.0 | 899 K | from 26.7 GB |
| 87 | MiniMax-M2.5 | 228.7 B | 197 K | other | 893 K | from 288.0 GB |
| 88 | Llama-3.2-1B-Instruct | 1.2 B | 131 K | ⚠ llama3.2 | 892 K | from 3.4 GB |
| 89 | tiny-gpt2 | — | — | unknown | 887 K | — |
| 90 | Gemma-4-31B-IT-NVFP4 | 20.9 B | — | other | 879 K | from 39.5 GB |
| 91 | glm-5.3-exlr8-k2-k3-k4 | — | 1.0 M | other | 874 K | from 921.8 GB |
| 92 | GLM-5.3-Flash-GGUF | 321.3 B | — | ✓ mit | 857 K | from 102.9 GB |
| 93 | GLM-4.5-Air | 110.5 B | 131 K | ✓ mit | 850 K | from 260.1 GB |
| 94 | Qwen2.5-3B-Instruct-AWQ | 3.4 B | 33 K | other | 841 K | from 4.0 GB |
| 95 | Hy3 | 298.8 B | 262 K | ✓ apache-2.0 | 837 K | from 702.7 GB |
| 96 | Qwen3-0.6B | 752 M | 41 K | ✓ apache-2.0 | 836 K | from 2.3 GB |
| 97 | Qwen3.6-27B-NVFP4 | 18.2 B | — | ✓ apache-2.0 | 827 K | from 27.3 GB |
| 98 | DeepSeek-V4-Flash | 158.1 B | 1.0 M | ✓ mit | 809 K | from 199.8 GB |
| 99 | gpt2-large | 810 M | — | ✓ mit | 801 K | from 4.2 GB |
| 100 | Qwen2.5-14B-Instruct | 14.8 B | 33 K | ✓ apache-2.0 | 799 K | from 35.2 GB |
VRAM figures are estimates for the smallest available quantization at 8K context — see /methodology. Downloads refresh nightly from the Hugging Face API; API prices (blended 3:1 input:output per 1M tokens) and Artificial Analysis indices from OpenRouter.