Which AI models can the AMD Radeon AI PRO R9700 run?
Every model we track, checked against the AMD Radeon AI PRO R9700 at 8K context. The largest that runs well is Mixtral 8x7B.
VRAM32 GBAMD
Bandwidth640GB/s
Runs well49/69models at Q4_K_M or better
Fastest model~341tok/s · Qwen3 Coder Next 80B-A3B
Every model on the AMD Radeon AI PRO R9700
Bars show memory used out of 32 GB. Pick a row for the full breakdown.
49 Runs well
NameMemory usedBest quantSpeed
Mixtral 8x7B47BQ4_K_M · 30 GB~49 tok/s
Qwen3.6 35B-A3B36BQ6_K · 30.3 GB~156 tok/s- Qwen2.5 32B33BQ6_K · 29.6 GB~14 tok/s
DeepSeek R1 Distill Qwen 32B33BQ6_K · 29.6 GB~14 tok/s- Qwen3 32B33BQ6_K · 29.6 GB~14 tok/s
- QwQ 32B33BQ6_K · 29.6 GB~14 tok/s
- Qwen2.5 Coder 32B33BQ6_K · 29.6 GB~14 tok/s
Nemotron 3 Nano 30B-A3B32BQ6_K · 26.6 GB~134 tok/s
Gemma 4 31B31BQ6_K · 27.4 GB~15 tok/s
GLM-4.7 Flash 30B-A3B31BQ6_K · 26.6 GB~156 tok/s- Qwen3 30B-A3B31BQ6_K · 26.4 GB~142 tok/s
- Qwen3 Coder 30B-A3B31BQ6_K · 26.4 GB~142 tok/s
- Qwen3.8 27B28BQ8_0 · 30.7 GB~13 tok/s
- Qwen3.6 27B28BQ8_0 · 30.7 GB~13 tok/s
- Gemma 3 27B27BQ8_0 · 30.8 GB~13 tok/s
- Gemma 2 27B27BQ8_0 · 31.8 GB~13 tok/s
- Gemma 4 26B-A4B26BQ8_0 · 28.3 GB~95 tok/s
- Mistral Small 3.2 24B24BQ8_0 · 27.4 GB~15 tok/s
- Mistral Small 24B24BQ8_0 · 27 GB~15 tok/s
- Devstral Small 24B24BQ8_0 · 27 GB~15 tok/s
- Magistral Small 24B24BQ8_0 · 27 GB~15 tok/s
gpt-oss 20B21BQ8_0 · 23 GB~100 tok/s- Qwen3 14B15BFP16 · 31.5 GB~13 tok/s
- Qwen2.5 Coder 14B15BFP16 · 31.8 GB~13 tok/s
- DeepSeek R1 Distill Qwen 14B15BFP16 · 31.8 GB~13 tok/s
- Qwen2.5 14B15BFP16 · 31.6 GB~13 tok/s
Phi-4 14B15BFP16 · 31.7 GB~13 tok/s- Phi-4 Reasoning 14B15BFP16 · 31.7 GB~13 tok/s
- Mistral Nemo 12B12BFP16 · 26.3 GB~16 tok/s
- Gemma 3 12B12BFP16 · 25.9 GB~16 tok/s
- Gemma 4 12B12BFP16 · 24.9 GB~16 tok/s
- Gemma 2 9B9.2BFP16 · 21.2 GB~21 tok/s
- Qwen3 8B8.2BFP16 · 18.2 GB~23 tok/s
- DeepSeek R1 0528 Qwen3 8B8.2BFP16 · 18.2 GB~23 tok/s
Granite 3.3 8B8.2BFP16 · 18.3 GB~24 tok/s
Llama 3.1 8B8BFP16 · 17.7 GB~24 tok/s- Qwen2.5 7B7.6BFP16 · 16.3 GB~25 tok/s
- Qwen2.5 Coder 7B7.6BFP16 · 16.3 GB~25 tok/s
- DeepSeek R1 Distill Qwen 7B7.6BFP16 · 16.3 GB~25 tok/s
- Mistral 7B v0.37.3BFP16 · 16.2 GB~26 tok/s
- Gemma 3 4B4.3BFP16 · 9.5 GB~45 tok/s
- Qwen3 4B4BFP16 · 9.8 GB~48 tok/s
- Phi-4 Mini 3.8B3.8BFP16 · 9.4 GB~50 tok/s
- Phi-3.5 Mini 3.8B3.8BFP16 · 11.5 GB~50 tok/s
- Llama 3.2 3B3.2BFP16 · 8 GB~60 tok/s
- Qwen3 1.7B2BFP16 · 5.6 GB~95 tok/s
- Llama 3.2 1B1.2BFP16 · 3.3 GB~155 tok/s
- Gemma 3 1B1BFP16 · 2.6 GB~192 tok/s
- Qwen3 0.6B0.8BFP16 · 3 GB~256 tok/s
5 Fits with heavy quantization
NameMemory usedBest quantSpeed
15 Too big
NameMemory usedBest quantSpeed
Kimi K3.1 Pro 2.8T-A104B2800Bneeds 1698.3 GB32 GB free- Qwen3.8 2.4T-A95B2446Bneeds 1484.3 GB32 GB free
- DeepSeek V4 Pro 1.6T-A49B1600Bneeds 971.2 GB32 GB free
- Mistral Large 4 1T-A52B1000Bneeds 607.5 GB32 GB free
- DeepSeek V4.1 Flash763Bneeds 463.4 GB32 GB free
- GLM-5.3753Bneeds 457.8 GB32 GB free
- GLM-5 744B-A40B744Bneeds 452.4 GB32 GB free
MiniMax M3 428B-A23B427Bneeds 260.5 GB32 GB free- GLM-5.3 Flash 320B-A18B321Bneeds 195.3 GB32 GB free
- DeepSeek V4 Flash 284B-A13B284Bneeds 173 GB32 GB free
- Qwen3 235B-A22B235Bneeds 144.7 GB32 GB free
- Qwen3.8 Flash Next180Bneeds 109.9 GB32 GB free
- gpt-oss 120B117Bneeds 71.7 GB32 GB free
- Llama 4 Scout 17B-16E109Bneeds 68.3 GB32 GB free
- GLM-4.5-Air 106B-A12B106Bneeds 66.4 GB32 GB free
Run it in the cloud
Need more than 32 GB? Rent a bigger GPU by the hour.
Rent a cloud GPU by the hour. No upfront cost, stop any time.
Apple M2 Pro (32GB) 32 GB