What GPU do you need to run MiniCPM-V 4.7 35B A3B?
35.21B params 256K max context VisionTool useReasoning
The smallest GPU memory that runs MiniCPM-V 4.7 35B A3B well is 24 GB (Q4_K_M, e.g. RTX Spark). 45 of 132 GPUs run it well and 2 more are a tight fit. It's a mixture-of-experts model, so all weights must fit in memory but speed depends only on the active experts.
VRAM needed per quantization
| Quantization | Weights | 16K context | 32K context | Full 256K |
|---|---|---|---|---|
| Q4_K_M | 19.7 GB | 20.0 GB | 20.3 GB | 24.6 GB |
| Q8_0 | 34.4 GB | 34.7 GB | 35.0 GB | 39.2 GB |
Weights plus fp16 KV cache. Leave ~1 GB headroom for the inference engine.
Cheapest GPUs that can run MiniCPM-V 4.7 35B A3B
Lowest launch MSRP among GPUs that run its smallest quantization (Q4_K_M) well, with at least 16K context.
As an Amazon Associate, I earn from qualifying purchases.
GPUs that can run MiniCPM-V 4.7 35B A3B
Best quantization that fits on each GPU, with the context and speed it gets. For chips with several memory sizes, the smallest that works is shown.
NVIDIA
- RTX 4000 Ada 20 GB Q4_K_M 15K context · ~134 tok/s tight fit
- RTX 3090 24 GB Q4_K_M 225K context · ~349 tok/s
- RTX 3090 Ti 24 GB Q4_K_M 225K context · ~376 tok/s
- RTX 4090 24 GB Q4_K_M 225K context · ~376 tok/s
- RTX 4500 Ada 24 GB Q4_K_M 225K context · ~161 tok/s
- RTX A5000 24 GB Q4_K_M 225K context · ~287 tok/s
- RTX PRO 4000 Blackwell 24 GB Q4_K_M 225K context · ~251 tok/s
- RTX Spark 24 GB Q4_K_M 225K context · ~112 tok/s
- Tesla P40 24 GB Q4_K_M 225K context · ~129 tok/s
- RTX 5090 32 GB Q4_K_M 256K context · ~669 tok/s
- RTX 5000 Ada 32 GB Q4_K_M 256K context · ~215 tok/s
- RTX PRO 4500 Blackwell 32 GB Q4_K_M 256K context · ~334 tok/s
- Tesla V100 32GB 32 GB Q4_K_M 256K context · ~336 tok/s
- L40S 48 GB Q8_0 256K context · ~220 tok/s
- RTX 6000 Ada 48 GB Q8_0 256K context · ~244 tok/s
- RTX A6000 48 GB Q8_0 256K context · ~195 tok/s
- RTX PRO 5000 Blackwell 48 GB Q8_0 256K context · ~342 tok/s
- RTX PRO 5000 Blackwell 72GB 72 GB Q8_0 256K context · ~342 tok/s
- RTX PRO 5500 Blackwell 84 GB Q8_0 256K context · ~356 tok/s
- RTX PRO 6000 Blackwell 96 GB Q8_0 256K context · ~456 tok/s
- DGX Spark (GB 10) 128 GB Q8_0 256K context · ~69 tok/s
AMD
- Radeon RX 7900 XT 20 GB Q4_K_M 15K context · ~299 tok/s tight fit
- Radeon RX 7900 XTX 24 GB Q4_K_M 225K context · ~358 tok/s
- Radeon AI PRO R9700 32 GB Q4_K_M 256K context · ~239 tok/s
- Ryzen AI Max (Strix Halo) 32 GB Q4_K_M 256K context · ~96 tok/s
- Ryzen AI Max 400 (Gorgon Halo) 64 GB Q8_0 256K context · ~69 tok/s
Intel
- Arc Pro B65 32 GB Q4_K_M 256K context · ~227 tok/s
- Arc Pro B70 32 GB Q4_K_M 256K context · ~227 tok/s
Apple
- M2 24 GB Q4_K_M 225K context · ~37 tok/s
- M3 24 GB Q4_K_M 225K context · ~37 tok/s
- M4 24 GB Q4_K_M 225K context · ~45 tok/s
- M4 Pro 24 GB Q4_K_M 225K context · ~102 tok/s
- M5 24 GB Q4_K_M 225K context · ~57 tok/s
- M5 Pro 24 GB Q4_K_M 225K context · ~115 tok/s
- M6 24 GB Q4_K_M 225K context · ~63 tok/s
- M1 Max 32 GB Q4_K_M 256K context · ~149 tok/s
- M1 Pro 32 GB Q4_K_M 256K context · ~75 tok/s
- M2 Max 32 GB Q4_K_M 256K context · ~149 tok/s
- M2 Pro 32 GB Q4_K_M 256K context · ~75 tok/s
- M3 Max 36 GB Q8_0 85K context · ~76 tok/s
- M3 Pro 36 GB Q8_0 85K context · ~38 tok/s
- M4 Max 36 GB Q8_0 85K context · ~104 tok/s
- M5 Max 36 GB Q8_0 85K context · ~117 tok/s
- M1 Ultra 64 GB Q8_0 256K context · ~204 tok/s
- M2 Ultra 64 GB Q8_0 256K context · ~204 tok/s
- M3 Ultra 96 GB Q8_0 256K context · ~208 tok/s
- M5 Ultra 96 GB Q8_0 256K context · ~305 tok/s