AMD Radeon PRO W7900
48 GB · 864 GB/s · 295 W
In short
The AMD Radeon PRO W7900 has 48 GB of GDDR6 memory with 864 GB/s of bandwidth, which sets how fast it generates text. At 32k context it runs 65 of the 125 open models we track fully in memory; the most capable is Qwen 3.8 27B at Q8_0, at about 15 tok/s (estimated). We have no current price for the AMD Radeon PRO W7900 in the US.
Updated . Speeds are estimates; how they are worked out.
Prices
Every retailer we track in the US, cheapest first. Buy links may earn us a commission.
No retailer in this market lists it right now.
What it runs
Every model in the catalogue, with the best quant for this hardware. Estimates, not measurements.
| Model | Quant | Fits | Memory | Max context | Decode @32k | Prefill | Good for |
|---|---|---|---|---|---|---|---|
| Qwen 3.8 27B Alibaba Qwen | Q8_0 | fits | 30.2 GB | 256k | 15tok/sest. | 692 tok/s | chat |
| Qwen 3.5 122B-A10B Alibaba Qwen | UD-Q2_K_XL | fits | 40.8 GB | 248k256k with q8 cache | 66tok/sest. | 1623 tok/s | chat, coding, agents, long documents |
| Qwen 3.6 27B Alibaba Qwen | Q8_0 | fits | 29.8 GB | 256k | 15tok/sest. | 692 tok/s | chat |
| Qwen3 Coder Next Alibaba Qwen | Q4_K_M | fits | 47.0 GB | 38k61k with q8 cache | 88tok/sest. | 3911 tok/s | chat, coding, agents, long documents |
| Gemma 4 31B | Q8_0 | fits | 34.7 GB | 181k256k with q8 cache | 13tok/sest. | 617 tok/s | – |
| Mistral Small 4 119B Mistral AI | UD-Q2_K_XL | fits | 39.1 GB | 321k511k with q8 cache | 91tok/sest. | 1804 tok/s | chat, coding, agents, long documents |
| Qwen 3.6 35B-A3B Alibaba Qwen | Q8_0 | fits | 36.0 GB | 256k | 75tok/sest. | 4941 tok/s | chat, coding, agents, long documents |
| Qwen 3.5 27B Alibaba Qwen | Q8_0 | fits | 29.8 GB | 256k | 15tok/sest. | 692 tok/s | chat |
| Gemma 4 26B-A4B | Q8_0 | fits | 26.8 GB | 256k | 57tok/sest. | 4396 tok/s | chat, coding, agents, long documents |
| Qwen3 Next 80B-A3B Alibaba Qwen | Q4_K_M | fits | 47.0 GB | 39k63k with q8 cache | 107tok/sest. | 4653 tok/s | chat, coding, agents, long documents |
| Qwen 3.5 35B-A3B Alibaba Qwen | Q8_0 | fits | 36.0 GB | 256k | 75tok/sest. | 4941 tok/s | chat, coding, agents, long documents |
| GLM 4.7 Flash Z.ai | Q8_0 | fits | 32.3 GB | 198k | 40tok/sest. | 2580 tok/s | chat, coding, agents, long documents |
| Devstral Small 2 24B Mistral AI | Q8_0 | fits | 29.3 GB | 142k259k with q8 cache | 16tok/sest. | 730 tok/s | chat |
| Nemotron 3.5 Lightning 30B-A3B NVIDIA | UD-Q6_K_XL | fits | 33.8 GB | 256k | 80tok/sest. | 5638 tok/s | chat, coding, agents, long documents |
| Kimi Linear 48B-A3B Moonshot AI | Q6_K | fits | 39.0 GB | 640k872k with q8 cache | 103tok/sest. | 4258 tok/s | chat, coding, long documents |
| gpt-oss 20B OpenAI | MXFP4 | fits | 13.0 GB | 128k | 88tok/sest. | 3689 tok/s | chat, coding, agents, long documents |
| Qwen3 Coder 30B-A3B Alibaba Qwen | Q8_0 | fits | 34.2 GB | 162k256k with q8 cache | 45tok/sest. | 2272 tok/s | chat, coding, agents, long documents |
| Qwen3 32B Alibaba Qwen | Q8_0 | fits | 41.4 GB | 40k | 11tok/sest. | 430 tok/s | – |
| Gemma 4 12B | Q8_0 | fits | 13.6 GB | 256k | 35tok/sest. | 1605 tok/s | chat, coding, long documents |
| Qwen 3.5 9B Alibaba Qwen | Q8_0 | fits | 10.9 GB | 256k | 45tok/sest. | 2000 tok/s | chat, coding, long documents |
| Mistral Small 3.2 24B Mistral AI | Q8_0 | fits | 29.3 GB | 128k | 16tok/sest. | 730 tok/s | chat |
| Gemma 3 27B | Q8_0 | fits | 30.6 GB | 128k | 15tok/sest. | 735 tok/s | chat |
| DeepSeek R1 Distill Qwen 32B DeepSeek | Q8_0 | fits | 41.4 GB | 54k100k with q8 cache | 11tok/sest. | 493 tok/s | – |
| Ministral 3 14B Mistral AI | Q8_0 | fits | 19.4 GB | 203k256k with q8 cache | 24tok/sest. | 1111 tok/s | chat, coding |
| Qwen3 14B Alibaba Qwen | Q8_0 | fits | 20.6 GB | 40k | 23tok/sest. | 999 tok/s | chat, coding |
| DeepSeek R1 0528 Qwen3 8B DeepSeek | Q8_0 | fits | 13.6 GB | 128k | 35tok/sest. | 1648 tok/s | chat, coding, long documents |
| Qwen3 8B Alibaba Qwen | Q8_0 | fits | 13.6 GB | 40k | 35tok/sest. | 1648 tok/s | chat, coding, long documents |
| Phi-4 14B Microsoft | Q8_0 | fits | 21.7 GB | 16k | 21tok/sest. | 1004 tok/s | chat, coding |
| Qwen 3.5 4B Alibaba Qwen | Q8_0 | fits | 6.2 GB | 256k | 85tok/sest. | 3742 tok/s | chat, coding, long documents |
| Ministral 3 8B Mistral AI | Q8_0 | fits | 13.6 GB | 256k | 35tok/sest. | 1591 tok/s | chat, coding, long documents |
| Granite 4.2 8B IBM | Q8_0 | fits | 14.7 GB | 128k | 32tok/sest. | 1515 tok/s | chat, coding, agents, long documents |
| Gemma 4 E4B | Q8_0 | fits | 9.5 GB | 128k | 52tok/sest. | 2513 tok/s | chat, coding, long documents |
| LFM2.5 8B-A1B Liquid AI | Q8_0 | fits | 10.9 GB | 125k | 107tok/sest. | 8218 tok/s | chat, coding, long documents |
| Llama 3.1 8B Meta | Q6_K | fits | 11.1 GB | 128k | 44tok/sest. | 1741 tok/s | chat, coding, long documents |
| Qwen2.5 0.5B Alibaba Qwen | Q8_0 | fits | 1.9 GB | 32k | 509tok/sest. | 17899 tok/s | chat, coding, long documents |
| Qwen3.5 0.8B Alibaba Qwen | Q8_0 | fits | 2.1 GB | 256k | 385tok/sest. | 16813 tok/s | chat, coding, agents, long documents |
| Qwen3.5 2B Alibaba Qwen | Q5_K_M | fits | 2.7 GB | 256k | 256tok/sest. | 8015 tok/s | chat, coding, agents, long documents |
| DeepSeek R1 Distill Qwen 1.5B DeepSeek | Q8_0 | fits | 3.6 GB | 128k | 168tok/sest. | 6733 tok/s | chat, coding, long documents |
| MiniCPM5 2B openbmb | Q8_0 | fits | 4.8 GB | 128k | 116tok/sest. | 4021 tok/s | chat, coding, agents, long documents |
| Spark X2.5 4B XHToken | mlx-4bit | fits | 5.0 GB | 1024k | 110tok/sest. | 3949 tok/s | chat, coding, agents, long documents |
| LFM2.5 2.6B Liquid AI | Q8_0 | fits | 5.5 GB | 128k | 97tok/sest. | 4555 tok/s | chat, coding, agents, long documents |
| LFM2.5 VL 3B Liquid AI | Q8_0 | fits | 5.5 GB | 32k | 97tok/sest. | 4177 tok/s | chat, coding, agents, long documents |
| Mellum2.1 12B A2.5B Thinking JetBrains | Q8_0 | fits | 13.5 GB | 128k | 90tok/sest. | 5934 tok/s | chat, coding, agents, long documents |
| FrogNano 4B 2609 Microsoft | Q8_0 | fits | 6.3 GB | 256k | 83tok/sest. | 3742 tok/s | chat, coding, agents, long documents |
| granite vision 4.1 4b IBM | Q8_0 | fits | 6.9 GB | 128k | 75tok/sest. | 2918 tok/s | chat, coding, long documents |
| granite 4.2 3b IBM | Q8_0 | fits | 7.1 GB | 128k | 72tok/sest. | 3058 tok/s | chat, coding, agents, long documents |
| Kimi VL A3B Thinking 2506 Moonshot AI | Q8_0 | fits | 17.7 GB | 128k | 69tok/sest. | 4459 tok/s | chat, coding, long documents |
| Kimi VL A3B Moonshot AI | mlx-8bit | fits | 18.5 GB | 128k | 67tok/sest. | 4459 tok/s | chat, coding, agents, long documents |
| Ministral 3 3B Instruct 2512 Mistral AI | Q8_0 | fits | 7.6 GB | 256k | 67tok/sest. | 2924 tok/s | chat, coding, agents, long documents |
| LFM2.5 24B A2B Liquid AI | Q8_0 | fits | 27.2 GB | 128k | 59tok/sest. | 4580 tok/s | chat, coding, agents, long documents |
| GLM 4 9B 0414 Z.ai | mlx-5bit | fits | 9.9 GB | 32k | 50tok/sest. | 1453 tok/s | chat, coding, agents, long documents |
| DeepSeek R1 Distill Qwen 7B DeepSeek | Q8_0 | fits | 10.3 GB | 128k | 48tok/sest. | 1968 tok/s | chat, coding, long documents |
| GLM 4.6V Flash Z.ai | Q8_0 | fits | 11.6 GB | 128k | 42tok/sest. | 1370 tok/s | chat, coding, agents, long documents |
| DeepSeek R1 Distill Llama 8B DeepSeek | Q8_0 | fits | 12.9 GB | 128k | 37tok/sest. | 1741 tok/s | chat, coding, long documents |
| GLM 4.1V 9B Thinking Z.ai | mlx-8bit | fits | 13.2 GB | 64k | 36tok/sest. | 1370 tok/s | chat, coding, long documents |
| Ministral 3 8B Reasoning 2512 Mistral AI | Q8_0 | fits | 13.6 GB | 256k | 35tok/sest. | 1591 tok/s | chat, coding, agents, long documents |
| Ministral 3 14B Reasoning 2512 Mistral AI | Q8_0 | fits | 19.4 GB | 203k256k with q8 cache | 24tok/sest. | 1111 tok/s | chat, coding |
| DeepSeek R1 Distill Qwen 14B DeepSeek | Q8_0 | fits | 21.6 GB | 128k | 22tok/sest. | 940 tok/s | chat, coding |
| Qwen3.8 27B OBLITERATED OBLITERATUS | mlx-4bit | fits | 22.4 GB | 256k | 21tok/sest. | 692 tok/s | chat, coding, long documents |
| Devstral Small 2507 Mistral AI | Q8_0 | fits | 29.3 GB | 128k | 16tok/sest. | 741 tok/s | chat |
| Magistral Small 2506 Mistral AI | Q8_0 | fits | 29.3 GB | 40k | 16tok/sest. | 741 tok/s | chat |
| Magistral Small 2507 Mistral AI | Q8_0 | fits | 29.3 GB | 40k | 16tok/sest. | 741 tok/s | chat |
| Magistral Small 2509 Mistral AI | Q8_0 | fits | 29.3 GB | 128k | 16tok/sest. | 730 tok/s | chat |
| Muse Glimmer 30B meta-models | Q8_0 | fits | 31.6 GB | 128k | 15tok/sest. | 667 tok/s | – |
| granite 4.2 30b IBM | Q8_0 | fits | 38.0 GB | 68k125k with q8 cache | 12tok/sest. | 567 tok/s | – |
| Nemotron 3 Super 120B-A12B NVIDIA | UD-Q2_K_XL | offload | 52.3 GB | – | 28tok/sest. | 1543 tok/s | chat, coding, long documents |
| gpt-oss 120B OpenAI | Q4_K_M | offload | 60.6 GB | – | 24tok/sest. | 2268 tok/s | chat, coding |
| GLM 4.5 Air Z.ai | Q2_K | offload | 48.9 GB | 22k41k with q8 cache | 19tok/sest. | 655 tok/s | chat |
| Llama 3.3 70B Meta | Q4_K_M | offload | 50.6 GB | 21k39k with q8 cache | 6tok/sest. | 224 tok/s | – |
| GLM 4.5V Z.ai | Q4_K_M | offload | 66.0 GB | – | 6tok/sest. | 541 tok/s | – |
| DeepSeek R1 Distill Llama 70B DeepSeek | Q4_K_M | offload | 50.6 GB | 21k39k with q8 cache | 6tok/sest. | 224 tok/s | – |
| GLM 4.6V Z.ai | Q4_K_M | offload | 72.3 GB | – | 5tok/sest. | 516 tok/s | – |
| Kimi Dev 72B Moonshot AI | Q6_K | offload | 70.9 GB | – | 2tok/sest. | 182 tok/s | – |
| DeepSeek V4 Flash DeepSeek | UD-Q2_K_XL | won't run | 93.9 GB | – | – | – | – |
| GLM 5.3 Flash Z.ai | UD-Q2_K_XL | won't run | 102.7 GB | – | – | – | – |
| Qwen 3.8 Flash Next Alibaba Qwen | UD-Q2_K_XL | won't run | 75.3 GB | – | – | – | – |
| MiniMax M2.7 MiniMax | UD-Q2_K_XL | won't run | 78.9 GB | – | – | – | – |
| DeepSeek R1 DeepSeek | Q4_K_M | won't run | 379.8 GB | – | – | – | – |
| DeepSeek R1 0528 DeepSeek | Q4_K_M | won't run | 380.3 GB | – | – | – | – |
| DeepSeek R1 Zero DeepSeek | Q4_K_M | won't run | 379.8 GB | – | – | – | – |
| DeepSeek V3 DeepSeek | Q4_K_M | won't run | 379.8 GB | – | – | – | – |
| DeepSeek V3 0324 DeepSeek | Q4_K_M | won't run | 379.8 GB | – | – | – | – |
| DeepSeek V3.1 DeepSeek | Q4_K_M | won't run | 380.7 GB | – | – | – | – |
| DeepSeek V3.1 Terminus DeepSeek | Q4_K_M | won't run | 380.7 GB | – | – | – | – |
| DeepSeek V3.2 DeepSeek | Q4_K_M | won't run | 380.7 GB | – | – | – | – |
| DeepSeek V4 Flash 0731 DeepSeek | MXFP4 | won't run | 149.3 GB | – | – | – | – |
| DeepSeek V4 Flash Vision Exp DeepSeek | MXFP4 | won't run | 158.1 GB | – | – | – | – |
| Devstral 2 123B Instruct 2512 Mistral AI | Q4_K_M | won't run | 81.7 GB | – | – | – | – |
| GLM 4.5 Z.ai | Q4_K_M | won't run | 214.1 GB | – | – | – | – |
| GLM 4.6 Z.ai | Q4_K_M | won't run | 213.3 GB | – | – | – | – |
| GLM 4.7 Z.ai | Q4_K_M | won't run | 214.1 GB | – | – | – | – |
| GLM 5 Z.ai | MXFP4 | won't run | 386.1 GB | – | – | – | – |
| GLM 5.1 Z.ai | MXFP4 | won't run | 423.7 GB | – | – | – | – |
| GLM 5.2 Z.ai | UD-Q4_K_M | won't run | 437.6 GB | – | – | – | – |
| GLM 5.3 Z.ai | Q8_0 | won't run | 750.1 GB | – | – | – | – |
| Kimi K2 Moonshot AI | Q4_K_M | won't run | 581.3 GB | – | – | – | – |
| Kimi K2 Instruct 0905 Moonshot AI | Q4_K_M | won't run | 581.7 GB | – | – | – | – |
| Kimi K2 Thinking Moonshot AI | Q4_K_M | won't run | 581.7 GB | – | – | – | – |
| Kimi K2.5 Moonshot AI | Q4_K_M | won't run | 581.7 GB | – | – | – | – |
| MiniMax M2 MiniMax | Q4_K_M | won't run | 137.6 GB | – | – | – | – |
| MiniMax M2.1 MiniMax | Q4_K_M | won't run | 137.6 GB | – | – | – | – |
| MiniMax M2.5 MiniMax | MXFP4 | won't run | 124.0 GB | – | – | – | – |
| MiniMax M3 MiniMax | MXFP4 | won't run | 244.4 GB | – | – | – | – |
| Mistral Medium 3.5 128B Mistral AI | Q4_K_M | won't run | 81.7 GB | – | – | – | – |
Specifications
From the manufacturer's published specifications; sources are linked where we have them.
- GPU
- Radeon PRO W7900
- Memory
- 48 GB GDDR6
- Memory bus
- 384-bit
- Memory bandwidth
- 864 GB/s
- FP16 compute
- 122.6 TFLOPS
- Software
- ROCm
- Board power
- 295 W
- PCIe generation
- Gen 4
- PCIe lanes
- x16
Compare with
Questions
- How much memory does the AMD Radeon PRO W7900 have for AI models?
- The AMD Radeon PRO W7900 has 48 GB of GDDR6 memory with 864 GB/s of bandwidth, which sets how fast it generates text.
- What is the best model the AMD Radeon PRO W7900 can run locally?
- Qwen 3.8 27B (Q8_0) fits fully in its memory at 32k context and generates at about 15 tok/s, an estimate from its memory bandwidth. 65 of the 125 models we track fit at that context.
- Can the AMD Radeon PRO W7900 run DeepSeek V4 Flash?
- No: even at UD-Q2_K_XL, DeepSeek V4 Flash needs about 94 GB at 32k context, more than the AMD Radeon PRO W7900 and system RAM can hold.
- How much does the AMD Radeon PRO W7900 cost in the US?
- We have no current price for the AMD Radeon PRO W7900 in the US.
Speeds on this page are for 32k context. See the best hardware for local AI at every budget.