readygpu

NVIDIA GeForce RTX 5090 vs NVIDIA RTX 6000 Ada Generation

For running models locally, priced in the US. At 32k context the first runs 61 models fully in memory, the second 65.

In short

The NVIDIA RTX 6000 Ada Generation has more memory for models: 48 GB against 32 GB. At 32k context the NVIDIA RTX 6000 Ada Generation runs 65 of the models we track fully in memory and the NVIDIA GeForce RTX 5090 61; only the NVIDIA RTX 6000 Ada Generation holds Qwen 3.5 122B-A10B. On the models both run fully in memory, the NVIDIA GeForce RTX 5090 is about 1.9× as fast as the NVIDIA RTX 6000 Ada Generation (estimated, median across models).

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

NVIDIA GeForce RTX 5090

$4,921

eBay

Add to build
NVIDIA RTX 6000 Ada Generation

Not priced here

Add to build

Specifications

SpecNVIDIA GeForce RTX 5090NVIDIA RTX 6000 Ada Generation
GPUGeForce RTX 5090RTX 6000 Ada Generation
Memory32 GB GDDR748 GB GDDR6
Memory bus512-bit384-bit
Memory bandwidth1,792 GB/s960 GB/s
FP16 compute209 TFLOPS91.1 TFLOPS
SoftwareCUDACUDA
Board power575 W300 W
PCIe generationGen 5Gen 4
PCIe lanesx16x16

What each runs

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

ModelNVIDIA GeForce RTX 5090NVIDIA RTX 6000 Ada Generation
Qwen 3.8 27B35tok/sQ8_019tok/sQ8_0
Qwen 3.5 122B-A10Boffload83tok/sUD-Q2_K_XL
Qwen 3.6 27B36tok/sQ8_019tok/sQ8_0
Gemma 4 31B39tok/sQ6_K17tok/sQ8_0
Qwen3 Coder Nextoffload110tok/sQ4_K_M
Mistral Small 4 119Boffload114tok/sUD-Q2_K_XL
Qwen 3.6 35B-A3B201tok/sQ6_K94tok/sQ8_0
Qwen 3.5 27B36tok/sQ8_019tok/sQ8_0
Gemma 4 26B-A4B132tok/sQ8_071tok/sQ8_0
Qwen 3.5 35B-A3B212tok/sQ6_K94tok/sQ8_0
GLM 4.7 Flash114tok/sQ6_K51tok/sQ8_0
Devstral Small 2 24B37tok/sQ8_020tok/sQ8_0
Qwen3 Next 80B-A3Boffload134tok/sQ4_K_M
Nemotron 3.5 Lightning 30B-A3B214tok/sUD-Q5_K_M100tok/sUD-Q6_K_XL
Kimi Linear 48B-A3B311tok/sQ4_K_M129tok/sQ6_K
gpt-oss 20B206tok/sMXFP4110tok/sMXFP4
Qwen3 Coder 30B-A3B117tok/sQ6_K56tok/sQ8_0
Qwen3 32B35tok/sQ5_K_M14tok/sQ8_0
Gemma 4 12B82tok/sQ8_044tok/sQ8_0
Qwen 3.5 9B104tok/sQ8_056tok/sQ8_0
Mistral Small 3.2 24B37tok/sQ8_020tok/sQ8_0
DeepSeek R1 Distill Qwen 32B35tok/sQ5_K_M14tok/sQ8_0
Gemma 3 27B35tok/sQ8_019tok/sQ8_0
Ministral 3 14B56tok/sQ8_030tok/sQ8_0
Qwen3 14B53tok/sQ8_028tok/sQ8_0
DeepSeek R1 0528 Qwen3 8B82tok/sQ8_044tok/sQ8_0
Qwen3 8B82tok/sQ8_044tok/sQ8_0
Phi-4 14B50tok/sQ8_027tok/sQ8_0
Qwen 3.5 4B198tok/sQ8_0106tok/sQ8_0
Ministral 3 8B82tok/sQ8_044tok/sQ8_0
Granite 4.2 8B76tok/sQ8_041tok/sQ8_0
Gemma 4 E4B121tok/sQ8_065tok/sQ8_0
LFM2.5 8B-A1B249tok/sQ8_0134tok/sQ8_0
Llama 3.1 8B102tok/sQ6_K55tok/sQ6_K
Qwen2.5 0.5B1190tok/sQ8_0637tok/sQ8_0
Qwen3.5 0.8B901tok/sQ8_0483tok/sQ8_0
Qwen3.5 2B598tok/sQ5_K_M321tok/sQ5_K_M
DeepSeek R1 Distill Qwen 1.5B392tok/sQ8_0210tok/sQ8_0
MiniCPM5 2B272tok/sQ8_0146tok/sQ8_0
Spark X2.5 4B258tok/smlx-4bit138tok/smlx-4bit
LFM2.5 2.6B227tok/sQ8_0122tok/sQ8_0
LFM2.5 VL 3B227tok/sQ8_0122tok/sQ8_0
Mellum2.1 12B A2.5B Thinking210tok/sQ8_0113tok/sQ8_0
FrogNano 4B 2609193tok/sQ8_0104tok/sQ8_0
granite vision 4.1 4b176tok/sQ8_094tok/sQ8_0
granite 4.2 3b169tok/sQ8_091tok/sQ8_0
Kimi VL A3B Thinking 2506162tok/sQ8_087tok/sQ8_0
Kimi VL A3B156tok/smlx-8bit84tok/smlx-8bit
Ministral 3 3B Instruct 2512156tok/sQ8_083tok/sQ8_0
LFM2.5 24B A2B139tok/sQ8_074tok/sQ8_0
GLM 4 9B 0414116tok/smlx-5bit62tok/smlx-5bit
DeepSeek R1 Distill Qwen 7B111tok/sQ8_060tok/sQ8_0
GLM 4.6V Flash98tok/sQ8_053tok/sQ8_0
DeepSeek R1 Distill Llama 8B87tok/sQ8_046tok/sQ8_0
GLM 4.1V 9B Thinking85tok/smlx-8bit45tok/smlx-8bit
Ministral 3 8B Reasoning 251282tok/sQ8_044tok/sQ8_0
Ministral 3 14B Reasoning 251256tok/sQ8_030tok/sQ8_0
DeepSeek R1 Distill Qwen 14B50tok/sQ8_027tok/sQ8_0
Qwen3.8 27B OBLITERATED48tok/smlx-4bit26tok/smlx-4bit
Muse Glimmer 30B46tok/sQ6_K18tok/sQ8_0
granite 4.2 30b38tok/sQ5_K_M15tok/sQ8_0
Devstral Small 250737tok/sQ8_020tok/sQ8_0
Magistral Small 250637tok/sQ8_020tok/sQ8_0
Magistral Small 250737tok/sQ8_020tok/sQ8_0
Magistral Small 250937tok/sQ8_020tok/sQ8_0
Nemotron 3 Super 120B-A12Boffloadoffload
gpt-oss 120Bwon't runoffload
GLM 4.5 Airoffloadoffload
Llama 3.3 70Boffloadoffload
DeepSeek R1 Distill Llama 70Boffloadoffload
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
MiniMax M2.7won'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 NVIDIA GeForce RTX 5090 or the NVIDIA RTX 6000 Ada Generation?
Choose the NVIDIA RTX 6000 Ada Generation to run bigger models. Choose the NVIDIA GeForce RTX 5090 for speed.
Which can run bigger models, the NVIDIA GeForce RTX 5090 or the NVIDIA RTX 6000 Ada Generation?
At 32k context the NVIDIA RTX 6000 Ada Generation runs 65 of the models we track fully in memory and the NVIDIA GeForce RTX 5090 61; only the NVIDIA RTX 6000 Ada Generation holds Qwen 3.5 122B-A10B.
Which is faster for local AI, the NVIDIA GeForce RTX 5090 or the NVIDIA RTX 6000 Ada Generation?
On the models both run fully in memory, the NVIDIA GeForce RTX 5090 is about 1.9× as fast as the NVIDIA RTX 6000 Ada Generation (estimated, median across models).

More comparisons

More hardware for local AI