VRAM.guide

Models GPUs Guides Method

Can I run

Index

Direct answers

One page per pairing, with the verdict at the top.

RTX 5090

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

RTX 4090

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

RTX 3090

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

RTX 5080

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

RTX 5070 Ti

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

RTX 5070

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

RTX 4070 Ti Super

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

RTX 4070 Super

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

RTX 4070

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

RTX 4060 Ti 16GB

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

RTX 3080 10GB

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

RTX 3060 12GB

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

RTX 5060 Ti 16GB

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

RTX 4080 Super

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

Radeon RX 7900 XTX

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

Arc B580

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

Mac Studio M4 Max 128GB

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

Mac Studio M3 Ultra 256GB

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

Mac Mini M4 Pro 48GB

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

NVIDIA DGX Spark (GB10)

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

Ryzen AI Max+ 395 128GB

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

A100 80GB

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

H100 SXM 80GB

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

L40S

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

RTX A6000

Llama 3.1 8B
8.0B Llama 3.1 70B
70.5B Llama 3.2 3B
3.2B Llama 3.2 1B
1.2B Llama 4 Scout 109B-A17B
109B Qwen3 8B
8.2B Qwen3 14B
14.8B Qwen3 32B
32.8B Qwen3 4B
4.0B Qwen3 30B-A3B
30.5B Qwen3 235B-A22B
235.1B Qwen2.5-Coder 32B
32.8B Qwen2.5 7B
7.6B Qwen2.5 72B
72.7B Gemma 3 4B
4.3B Gemma 3 12B
12.2B Gemma 3 27B
27.4B Mistral 7B v0.3
7.3B Mistral NeMo 12B
12.3B Mistral Small 24B
23.6B Mixtral 8x7B
46.7B DeepSeek-R1 671B-A37B
671B DeepSeek-R1-Distill-Qwen 32B
32.8B DeepSeek-R1-Distill-Llama 8B
8.0B gpt-oss 20B-A3.6B
20.9B gpt-oss 120B-A5.1B
116.8B Phi-4 14B
14.7B GLM-4 9B
9.4B Command R 35B
35.0B Yi-1.5 34B
34.4B

VRAM Guide — estimates, not guarantees. How the numbers are worked out. Report a wrong number