readygpu

AMD Radeon AI PRO R9700 vs AMD Radeon PRO W7900

For running models locally, priced in the US. At 8k context the first runs 61 models fully in memory, the second 68.

In short

The AMD Radeon PRO W7900 has more memory for models: 48 GB against 32 GB. At 8k context the AMD Radeon PRO W7900 runs 68 of the models we track fully in memory and the AMD Radeon AI PRO R9700 61; only the AMD Radeon PRO W7900 holds Qwen 3.5 122B-A10B. On the models both run fully in memory, the AMD Radeon PRO W7900 is about 1.3× as fast as the AMD Radeon AI PRO R9700 (estimated, median across models).

Updated . Speeds are estimates; how they are worked out.

AMD Radeon AI PRO R9700

$1,944

Amazon

Add to build
AMD Radeon PRO W7900

Not priced here

Add to build

Specifications

SpecAMD Radeon AI PRO R9700AMD Radeon PRO W7900
GPURadeon AI PRO R9700Radeon PRO W7900
Memory32 GB GDDR648 GB GDDR6
Memory bus256-bit384-bit
Memory bandwidth640 GB/s864 GB/s
FP16 compute191 TFLOPS122.6 TFLOPS
SoftwareROCmROCm
Board power300 W295 W
PCIe generationGen 5Gen 4
PCIe lanesx16x16

What each runs

Estimated decode speed with the most precise quant that fits each one est.

ModelAMD Radeon AI PRO R9700AMD Radeon PRO W7900
Qwen 3.8 27B12tok/sQ8_016tok/sQ8_0
Qwen 3.5 122B-A10Boffload77tok/sUD-Q2_K_XL
Qwen 3.6 27B12tok/sQ8_016tok/sQ8_0
Gemma 4 31B13tok/sQ6_K14tok/sQ8_0
Qwen3 Coder Nextoffload108tok/sQ4_K_M
Mistral Small 4 119Boffload111tok/sUD-Q2_K_XL
Qwen 3.6 35B-A3B75tok/sQ6_K86tok/sQ8_0
Qwen 3.5 27B12tok/sQ8_016tok/sQ8_0
Gemma 4 26B-A4B47tok/sQ8_063tok/sQ8_0
GLM 4.5 Airoffload39tok/sQ2_K
Qwen 3.5 35B-A3B80tok/sQ6_K86tok/sQ8_0
GLM 4.7 Flash37tok/sQ8_050tok/sQ8_0
Devstral Small 2 24B13tok/sQ8_018tok/sQ8_0
Qwen3 Next 80B-A3Boffload138tok/sQ4_K_M
Nemotron 3.5 Lightning 30B-A3B71tok/sUD-Q5_K_M83tok/sUD-Q6_K_XL
Kimi Linear 48B-A3B109tok/sQ4_K_M111tok/sQ6_K
gpt-oss 20B80tok/sMXFP4108tok/sMXFP4
Qwen3 Coder 30B-A3B65tok/sQ6_K71tok/sQ8_0
Qwen3 32B12tok/sQ6_K13tok/sQ8_0
Gemma 4 12B27tok/sQ8_036tok/sQ8_0
Qwen 3.5 9B36tok/sQ8_048tok/sQ8_0
Mistral Small 3.2 24B13tok/sQ8_018tok/sQ8_0
DeepSeek R1 Distill Qwen 32B12tok/sQ6_K13tok/sQ8_0
Gemma 3 27B12tok/sQ8_016tok/sQ8_0
Llama 3.3 70Boffload11tok/sQ4_K_M
Ministral 3 14B22tok/sQ8_030tok/sQ8_0
Qwen3 14B21tok/sQ8_028tok/sQ8_0
DeepSeek R1 0528 Qwen3 8B36tok/sQ8_048tok/sQ8_0
Qwen3 8B36tok/sQ8_048tok/sQ8_0
Phi-4 14B20tok/sQ8_028tok/sQ8_0
Qwen 3.5 4B73tok/sQ8_099tok/sQ8_0
Ministral 3 8B35tok/sQ8_047tok/sQ8_0
Granite 4.2 8B33tok/sQ8_045tok/sQ8_0
Gemma 4 E4B42tok/sQ8_056tok/sQ8_0
LFM2.5 8B-A1B144tok/sQ8_0194tok/sQ8_0
Llama 3.1 8B46tok/sQ6_K62tok/sQ6_K
Qwen2.5 0.5B557tok/sQ8_0752tok/sQ8_0
Qwen3.5 0.8B378tok/sQ8_0510tok/sQ8_0
Qwen3.5 2B226tok/sQ5_K_M306tok/sQ5_K_M
DeepSeek R1 Distill Qwen 1.5B165tok/sQ8_0223tok/sQ8_0
MiniCPM5 2B116tok/sQ8_0157tok/sQ8_0
LFM2.5 2.6B104tok/sQ8_0141tok/sQ8_0
LFM2.5 VL 3B104tok/sQ8_0141tok/sQ8_0
Spark X2.5 4B104tok/smlx-4bit140tok/smlx-4bit
granite vision 4.1 4b82tok/sQ8_0111tok/sQ8_0
Ministral 3 3B Instruct 251278tok/sQ8_0105tok/sQ8_0
granite 4.2 3b77tok/sQ8_0104tok/sQ8_0
LFM2.5 24B A2B76tok/sQ8_0102tok/sQ8_0
Mellum2.1 12B A2.5B Thinking75tok/sQ8_0101tok/sQ8_0
FrogNano 4B 260971tok/sQ8_096tok/sQ8_0
Kimi VL A3B Thinking 250663tok/sQ8_085tok/sQ8_0
Kimi VL A3B60tok/smlx-8bit81tok/smlx-8bit
DeepSeek R1 Distill Qwen 7B41tok/sQ8_056tok/sQ8_0
GLM 4 9B 041441tok/smlx-5bit55tok/smlx-5bit
DeepSeek R1 Distill Llama 8B37tok/sQ8_049tok/sQ8_0
Ministral 3 8B Reasoning 251235tok/sQ8_047tok/sQ8_0
GLM 4.6V Flash34tok/sQ8_046tok/sQ8_0
GLM 4.1V 9B Thinking29tok/smlx-8bit39tok/smlx-8bit
Ministral 3 14B Reasoning 251222tok/sQ8_030tok/sQ8_0
DeepSeek R1 Distill Qwen 14B20tok/sQ8_027tok/sQ8_0
Qwen3.8 27B OBLITERATED17tok/smlx-4bit22tok/smlx-4bit
Devstral Small 250713tok/sQ8_018tok/sQ8_0
Magistral Small 250613tok/sQ8_018tok/sQ8_0
Magistral Small 250713tok/sQ8_018tok/sQ8_0
Magistral Small 250913tok/sQ8_018tok/sQ8_0
granite 4.2 30b13tok/sQ6_K14tok/sQ8_0
Muse Glimmer 30B11tok/sQ8_015tok/sQ8_0
DeepSeek R1 Distill Llama 70Boffload11tok/sQ4_K_M
MiniMax M2.7won't runoffload
Nemotron 3 Super 120B-A12Boffloadoffload
gpt-oss 120Bwon't runoffload
GLM 4.5Vwon't runoffload
GLM 4.6Vwon't runoffload
Kimi Dev 72Bwon't runoffload
DeepSeek V4 Flashwon't runwon't run
GLM 5.3 Flashwon't runwon't run
Qwen 3.8 Flash Nextwon't runwon't run
DeepSeek R1won't runwon't run
DeepSeek R1 0528won't runwon't run
DeepSeek R1 Zerowon't runwon't run
DeepSeek V3won't runwon't run
DeepSeek V3 0324won't runwon't run
DeepSeek V3.1won't runwon't run
DeepSeek V3.1 Terminuswon't runwon't run
DeepSeek V3.2won't runwon't run
DeepSeek V4 Flash 0731won't runwon't run
DeepSeek V4 Flash Vision Expwon't runwon't run
Devstral 2 123B Instruct 2512won't runwon't run
GLM 4.5won't runwon't run
GLM 4.6won't runwon't run
GLM 4.7won't runwon't run
GLM 5won't runwon't run
GLM 5.1won't runwon't run
GLM 5.2won't runwon't run
GLM 5.3won't runwon't run
Kimi K2won't runwon't run
Kimi K2 Instruct 0905won't runwon't run
Kimi K2 Thinkingwon't runwon't run
Kimi K2.5won't runwon't run
MiniMax M2won't runwon't run
MiniMax M2.1won't runwon't run
MiniMax M2.5won't runwon't run
MiniMax M3won't runwon't run
Mistral Medium 3.5 128Bwon't runwon't run

Questions

Which is better for local AI, the AMD Radeon AI PRO R9700 or the AMD Radeon PRO W7900?
Choose the AMD Radeon PRO W7900 to run bigger models. Choose the AMD Radeon PRO W7900 for speed.
Which can run bigger models, the AMD Radeon AI PRO R9700 or the AMD Radeon PRO W7900?
At 8k context the AMD Radeon PRO W7900 runs 68 of the models we track fully in memory and the AMD Radeon AI PRO R9700 61; only the AMD Radeon PRO W7900 holds Qwen 3.5 122B-A10B.
Which is faster for local AI, the AMD Radeon AI PRO R9700 or the AMD Radeon PRO W7900?
On the models both run fully in memory, the AMD Radeon PRO W7900 is about 1.3× as fast as the AMD Radeon AI PRO R9700 (estimated, median across models).

More comparisons

More hardware for local AI