What can NVIDIA RTX PRO 6000 Blackwell run for vision?

Build: RTX PRO 6000 Blackwell + Threadripper PRO + 128GB

Memory: 96 GB VRAM + 128 GB system RAM
Runner: llama.cpp / Ollama (CUDA)

Runs comfortably
37 models

Ranked by fit for vision use case + predicted speed. Click a row for VRAM breakdown.

#1Gemma 4 12B
12B
gemma
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 15.8 GBHeadroom: 80.2 GB
ollama run gemma4:12b
161
tok/s
Estimated
Weights
7.60 GB
KV cache
6.00 GB
Activations
0.39 GB
Runtime
1.80 GB
#2Qwen3.5 9B
9B
qwen
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 13.2 GBHeadroom: 82.8 GB
ollama run qwen3.5:9b
214
tok/s
Estimated
Weights
6.60 GB
KV cache
4.50 GB
Activations
0.34 GB
Runtime
1.80 GB
#3Gemma 4 E4B (Effective 4B)
4B
gemma
Commercial OK
Quant: Q8_0Context: 8,192VRAM: 8.4 GBHeadroom: 87.6 GB
ollama run gemma4:e4b
274
tok/s
Estimated
Weights
4.40 GB
KV cache
2.00 GB
Activations
0.23 GB
Runtime
1.80 GB
#4Ministral 3 14B
14B
mistral
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 18.4 GBHeadroom: 77.6 GB
ollama run ministral-3:14b
138
tok/s
Estimated
Weights
9.10 GB
KV cache
7.00 GB
Activations
0.46 GB
Runtime
1.80 GB
#5Gemma 3 4B
4B
gemma
Commercial OK
Quant: Q8_0Context: 8,192VRAM: 8.4 GBHeadroom: 87.6 GB
ollama run gemma3:4b
274
tok/s
Estimated
Weights
4.40 GB
KV cache
2.00 GB
Activations
0.23 GB
Runtime
1.80 GB
#6Gemma 4 E2B (Effective 2B)
2B
gemma
Commercial OK
Quant: Q8_0Context: 8,192VRAM: 5.1 GBHeadroom: 90.9 GB
ollama run gemma4:e2b
548
tok/s
Estimated
Weights
2.20 GB
KV cache
1.00 GB
Activations
0.12 GB
Runtime
1.80 GB
#7Phi-3.5 Vision
4.2B
phi
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 6.5 GBHeadroom: 89.5 GB
459
tok/s
Estimated
Weights
2.50 GB
KV cache
2.10 GB
Activations
0.13 GB
Runtime
1.80 GB
#8PaliGemma 2 3B
3B
gemma
Commercial OK
Quant: BF16Context: 8,192VRAM: 9.6 GBHeadroom: 86.4 GB
194
tok/s
Estimated
Weights
6.00 GB
KV cache
1.50 GB
Activations
0.31 GB
Runtime
1.80 GB
#9LLaVA 1.6 Mistral 7B
7B
other
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 10.0 GBHeadroom: 86.0 GB
276
tok/s
Estimated
Weights
4.50 GB
KV cache
3.50 GB
Activations
0.23 GB
Runtime
1.80 GB
#10Phi-4 Multimodal
14B
phi
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 18.3 GBHeadroom: 77.7 GB
138
tok/s
Estimated
Weights
9.00 GB
KV cache
7.00 GB
Activations
0.46 GB
Runtime
1.80 GB
#11LLaVA-OneVision 7B
7B
other
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 10.0 GBHeadroom: 86.0 GB
276
tok/s
Estimated
Weights
4.50 GB
KV cache
3.50 GB
Activations
0.23 GB
Runtime
1.80 GB
#12Moondream 2
1.9B
other
Commercial OK
Quant: Q4_K_MContext: 2,048VRAM: 3.3 GBHeadroom: 92.7 GB
1015
tok/s
Estimated
Weights
1.20 GB
KV cache
0.24 GB
Activations
0.06 GB
Runtime
1.80 GB

What if you upgraded?

Hypothetical scenarios. We re-ran the compatibility engine for each.

+32 GB system RAM

~$80–150

Doubles your CPU-offload working set. Helps when models don't quite fit in VRAM.

Unlocks: 227 new comfortable, 14 new tradeoff

  • Qwen 3 0.6B
  • Llama 4 Scout
  • Llama 3.1 8B Instruct
  • Qwen 3 30B-A3B

Upgrade to NVIDIA H200 NVL (PCIe)

~$32000

141 GB VRAM (vs your 96 GB) plus a bandwidth jump from ~1792 GB/s to ~4800 GB/s.

Unlocks: 232 new comfortable

  • Qwen 3 0.6B
  • Llama 4 Scout
  • Llama 3.1 8B Instruct
  • Qwen 3 30B-A3B

Add a second NVIDIA RTX PRO 6000 Blackwell

~$8999

Tensor parallelism splits the model across both cards, effectively doubling VRAM. Bandwidth doesn't double — runs ~1.5× the single-card speed in practice.

Unlocks: 241 new comfortable

  • Qwen 3 235B-A22B
  • Qwen 3 0.6B
  • Llama 4 Scout
  • Llama 3.1 8B Instruct

Some links above are affiliate links. We may earn a commission at no extra cost to you. How we make money.

Won't run
top 5 popular models

Need more memory than you have. Shown for orientation.

DeepSeek V4 Pro (1.6T MoE)
1600B
deepseek
Commercial OK

Even with CPU offload, needs more memory than your VRAM (96 GB) + 60% of system RAM (77 GB) combined.

Qwen 3.5 235B-A17B (MoE)
397B
qwen
Commercial OK

Even with CPU offload, needs more memory than your VRAM (96 GB) + 60% of system RAM (77 GB) combined.

Qwen 3 235B-A22B
235B
qwen
Commercial OK

Even with CPU offload, needs more memory than your VRAM (96 GB) + 60% of system RAM (77 GB) combined.

DeepSeek R1 (671B reasoning)
671B
deepseek
Commercial OK

Even with CPU offload, needs more memory than your VRAM (96 GB) + 60% of system RAM (77 GB) combined.

DeepSeek V4 Flash (284B MoE)
284B
deepseek
Commercial OK

Even with CPU offload, needs more memory than your VRAM (96 GB) + 60% of system RAM (77 GB) combined.

How to read these numbers

Measured here
Measured here - RunLocalAI ran this exact combo on owner hardware with public evidence.

Source-backed
Source-backed / community - a reproduced public source supports the speed, but it is not labeled as owner-measured.

Extrapolated
Extrapolated - predicted from a measured benchmark on similar-bandwidth hardware.

Estimated
Estimated - formula based on VRAM bandwidth and model architecture; not a benchmark row.

RunLocalAI Will-It-Run Framework →

Want a specific benchmark we don't have? Email Contact support and we'll prioritize it.