Compatibility Matrix
Best community run per VRAM tier for the most-run open-weight models. Cells show tok/s and quantization — the same model often lands in multiple tiers via different quants. Full board · Submit your run.
Model familyAllArceeBaichuanBTLCohereDeepseekDeepseek-CoderDeepseek-R1DeepSeek-V4.1-FlashGemmaGLMGptHunyuanKimiKwaipilotLagunaLFMLingLlamaMiniCPM-VMinimaxMistralMobileLLMMuseNemotronOrnithPhiQwenQwen3.8RNJSparkunknown
| Model | Family | <24GB | 24–48GB | >48GB | runs |
|---|---|---|---|---|---|
| Qwen3.8-27B | Qwen | 391.0 · NVFP4 | 4546.8 · NVFP4 | 1392.4 · W4A16 | 67 |
| Qwen3.8-Flash-Next | Qwen | 81.8 · Q2_0 | 145.8 · W4A16+FP8-KV | 204.5 · FP8_E4M3 | 40 |
| DeepSeek-V4.1-Flash | Deepseek | — | 26.0 · EXL3 3.0 bpw | 231.6 · FP4+FP8 native | 39 |
| GLM-5.3-Flash | GLM | — | 250.6 · compressed-tensors | 264.6 · W4A16 | 21 |
| Qwen3.6-35B-A3B | Qwen | 129.4 · UD-Q4_K_XL | 969.4 · MQ4R | 330.2 · NVFP4 | 17 |
| Muse-Glimmer-30B | Muse | 56.0 · UD-Q4_K_XL | 161.9 · UD-Q8_K_XL | 152.2 · NVFP4 | 13 |
| Qwen3.6-27B | Qwen | 60.6 · IQ4_XS | 240.8 · AutoRound INT4 | 101.4 · FP8 | 11 |
| Ornith-1.5-35B-A3B | Qwen | 64.4 · Q4_K_M | 258.8 · NVFP4 | 87.1 · Q8_0 | 9 |
| NVIDIA-Nemotron-3.5-Lightning-30B-A3B | Nemotron | 116.0 · NVFP4 | 446.7 · NVFP4 | 217.9 · W4A16 | 9 |
| Qwen 3.8 Flash | Qwen | — | — | 86.8 · EXL3 3.05bpw | 8 |
| Ornith-1.5-9B | Qwen | 102.0 · Q4_K_M | 49.6 · Q8_0 | 61.5 · BF16 | 8 |
| DeepSeek-V4-Flash-0731 | Deepseek | — | 91.2 · FP8_E4M3 | 185.6 · FP8 | 7 |
| Qwen3.8-27B-MTP | Qwen | — | 117.7 · NVFP4 | 112.1 · NVFP4 | 7 |
| Qwen3.8-Flash-Next-hibrid48 | Qwen | — | — | 474.0 · NVFP4 (output head) | 5 |
| LFM2.5-350M | LFM | 437.7 · MQ4 | 466.0 · MQ4 | — | 5 |
| Laguna-XS-2.1 | Laguna | — | 283.1 · Q5_K_M GGUF | — | 5 |
| Qwen3.5-0.8B | Qwen | 545.3 · MQ4 | 554.5 · MQ4 | — | 5 |
| gemma-4-26B-A4B-it | Gemma | 62.1 · UD-Q4_K_M | 209.8 · Q4_K_M | — | 4 |
| Qwen3.8-27B-Uncensored | Qwen | 74.1 · IQ2_XXS | 136.4 · GPTQ-Int4 | — | 4 |
| Ling-3.0-tiny | Ling | 52.6 · Q4_K_M | 232.8 · Q8_0 | 369.1 · Q4_K_M | 4 |
| gemma-4-12B-it | Gemma | 89.7 · Q4_K_XL | 88.3 · Q8_0 | — | 4 |
| Qwen3.6-27B-MTP | Qwen | — | 41.2 · Q4_K_M | 60.1 · Q6_K | 4 |
| gemma-4-31B-it | Gemma | 15.8 · Q4_K_M | 25.4 · Q4_K_XL | 28.3 · NVFP4 | 4 |
| Qwen3.8-9B-Distill | Qwen | 76.4 · Q4_K_M | — | — | 3 |
| Laguna-S-2.1 | Laguna | — | 125.5 · INT4_W4A16 | 63.3 · Q4_K_M | 3 |
| Qwen3-14B | Qwen | 75.0 · Q4_K_M | — | 90.8 · FP8 | 3 |
| NVIDIA-Nemotron-3-Nano-4B | Nemotron | 185.8 · Q4_K_M | 237.8 · Q8_0 | — | 3 |
| gpt-oss-20b | Gpt | 93.4 · Q4_K_M | 307.2 · MXFP4 | — | 3 |
| DeepSeek-V4.1-Flash-1017 | DeepSeek-V4.1-Flash | — | — | 190.0 · EXL3 3.5bpw | 3 |
| gemma-4-E2B-it | Gemma | 113.4 · Q4_K_M | 232.6 · Q4_K_M | — | 3 |
| gemma-4-E4B-it | Gemma | 70.9 · Q4_K_M | 161.5 · Q8_0 | — | 3 |
| gemma-4-12B-it-qat | Gemma | 71.2 · Q4_0 | 130.8 · Q4_0 | — | 3 |
| LFM2.5-230M | LFM | 481.8 · MQ4 | 520.0 · MQ4 | — | 3 |
| Qwen3-30B-A3B-Instruct-2507 | Qwen | — | 312.9 · Q4_K_M | — | 2 |
| Qwen3.5-9B | Qwen | 51.9 · Q4_K_M | 38.6 · bfp_bf4 | — | 2 |
| DeepSeek-R1-Distill-Qwen-7B | Deepseek-R1 | 39.8 · Q8_0 | 213.3 · Q4_K_M | — | 2 |
| Qwen3-0.6B | Qwen | 310.4 · Q8_0 | — | 956.4 · Q4_K_M | 2 |
| gemma-4-26B-A4B-it-qat | Gemma | 95.3 · Q4_K_M | 254.3 · Q4_0 | — | 2 |
| MiniMax-M2.7 | Minimax | — | 94.5 · INT4 AutoRound W4A16 | 495.9 · FP8 | 2 |
| Qwen3-Coder-30B-A3B-Instruct | Qwen | — | 263.0 · Q6_K | 26.1 · FP8 | 2 |