NVIDIA GeForce RTX 3090 vs NVIDIA RTX PRO 6000 Blackwell

For running models locally, priced in Canada. At 8k context the first runs 28 models fully in memory, the second 41.

In short

The NVIDIA RTX PRO 6000 Blackwell has more memory for models: 96 GB against 24 GB. At 8k context the NVIDIA RTX PRO 6000 Blackwell runs 41 of the models we track fully in memory and the NVIDIA GeForce RTX 3090 28; only the NVIDIA RTX PRO 6000 Blackwell holds DeepSeek V4 Flash. On the models both run fully in memory, the NVIDIA RTX PRO 6000 Blackwell is about 1.8× as fast as the NVIDIA GeForce RTX 3090 (estimated, median across models). In Canada the NVIDIA GeForce RTX 3090 costs $2,800 at Amazon and the NVIDIA RTX PRO 6000 Blackwell $3,133 at Amazon: the NVIDIA GeForce RTX 3090 is $333 cheaper.

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

Specifications

SpecNVIDIA GeForce RTX 3090NVIDIA RTX PRO 6000 Blackwell
GPUGeForce RTX 3090RTX PRO 6000 Blackwell
Memory24 GB GDDR6X96 GB GDDR7
Memory bus384-bit512-bit
Memory bandwidth936 GB/s1,792 GB/s
FP16 compute71 TFLOPS504 TFLOPS
SoftwareCUDACUDA
Board power350 W600 W
PCIe generationGen 4Gen 5
PCIe lanesx16x16

What each runs

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

ModelNVIDIA GeForce RTX 3090NVIDIA RTX PRO 6000 Blackwell
DeepSeek V4 Flashwon't run89tok/sUD-Q2_K_XL
Qwen 3.8 Flash Nextwon't run21tok/sUD-Q3_K_XL
Qwen 3.8 27B26tok/sUD-Q6_K37tok/sQ8_0
MiniMax M2.7won't run117tok/sUD-Q2_K_XL
Qwen 3.5 122B-A10Boffload79tok/sQ6_K
Qwen 3.6 27B25tok/sQ6_K38tok/sQ8_0
Gemma 4 31B25tok/sQ5_K_M33tok/sQ8_0
Qwen3 Coder Nextoffload151tok/sQ8_0
Nemotron 3 Super 120B-A12Bwon't run81tok/sUD-Q4_K_M
gpt-oss 120Bwon't run195tok/sMXFP4
Mistral Small 4 119Boffload110tok/sUD-Q6_K
Qwen 3.6 35B-A3B168tok/sUD-Q4_K_M201tok/sQ8_0
Qwen 3.5 27B25tok/sQ6_K38tok/sQ8_0
Gemma 4 26B-A4B88tok/sUD-Q6_K147tok/sQ8_0
GLM 4.5 Airoffload40tok/sQ6_K
Qwen 3.5 35B-A3B168tok/sQ4_K_M201tok/sQ8_0
GLM 4.7 Flash87tok/sQ5_K_M117tok/sQ8_0
Devstral Small 2 24B28tok/sQ6_K42tok/sQ8_0
Qwen3 Next 80B-A3Boffload196tok/sQ8_0
Nemotron 3.5 Lightning 30B-A3Boffload195tok/sUD-Q6_K_XL
gpt-oss 20B132tok/sMXFP4253tok/sMXFP4
Qwen3 Coder 30B-A3B118tok/sQ5_K_M167tok/sQ8_0
Kimi Linear 48B-A3Boffload202tok/sQ8_0
Qwen3 32B27tok/sQ4_K_M30tok/sQ8_0
Gemma 4 12B44tok/sQ8_085tok/sQ8_0
Qwen 3.5 9B59tok/sQ8_0113tok/sQ8_0
Mistral Small 3.2 24B28tok/sQ6_K42tok/sQ8_0
DeepSeek R1 Distill Qwen 32B26tok/sQ4_K_M30tok/sQ8_0
Gemma 3 27B25tok/sQ6_K37tok/sQ8_0
Llama 3.3 70Boffload14tok/sQ8_0
Ministral 3 14B37tok/sQ8_071tok/sQ8_0
Qwen3 14B34tok/sQ8_065tok/sQ8_0
DeepSeek R1 0528 Qwen3 8B59tok/sQ8_0112tok/sQ8_0
Qwen3 8B59tok/sQ8_0112tok/sQ8_0
Phi-4 14B34tok/sQ8_064tok/sQ8_0
Qwen 3.5 4B121tok/sQ8_0231tok/sQ8_0
Ministral 3 8B57tok/sQ8_0109tok/sQ8_0
Granite 4.2 8B54tok/sQ8_0104tok/sQ8_0
Gemma 4 E4B69tok/sQ8_0131tok/sQ8_0
LFM2.5 8B-A1B237tok/sQ8_0455tok/sQ8_0
Llama 3.1 8B76tok/sQ6_K145tok/sQ6_K
GLM 5.3 Flashwon't runoffload

Questions

Which is better for local AI, the NVIDIA GeForce RTX 3090 or the NVIDIA RTX PRO 6000 Blackwell?
Choose the NVIDIA RTX PRO 6000 Blackwell to run bigger models. Choose the NVIDIA RTX PRO 6000 Blackwell for speed. Choose the NVIDIA GeForce RTX 3090 to spend less.
Which can run bigger models, the NVIDIA GeForce RTX 3090 or the NVIDIA RTX PRO 6000 Blackwell?
At 8k context the NVIDIA RTX PRO 6000 Blackwell runs 41 of the models we track fully in memory and the NVIDIA GeForce RTX 3090 28; only the NVIDIA RTX PRO 6000 Blackwell holds DeepSeek V4 Flash.
Which is faster for local AI, the NVIDIA GeForce RTX 3090 or the NVIDIA RTX PRO 6000 Blackwell?
On the models both run fully in memory, the NVIDIA RTX PRO 6000 Blackwell is about 1.8× as fast as the NVIDIA GeForce RTX 3090 (estimated, median across models).
What is the price difference between the NVIDIA GeForce RTX 3090 and the NVIDIA RTX PRO 6000 Blackwell?
In Canada the NVIDIA GeForce RTX 3090 costs $2,800 at Amazon and the NVIDIA RTX PRO 6000 Blackwell $3,133 at Amazon: the NVIDIA GeForce RTX 3090 is $333 cheaper.

More comparisons