readygpu

NVIDIA GeForce RTX 5070 Ti vs NVIDIA RTX A6000

For running models locally, priced in the US. At 128k context the first runs 22 models fully in memory, the second 60.

In short

The NVIDIA RTX A6000 has more memory for models: 48 GB against 16 GB. At 128k context the NVIDIA RTX A6000 runs 60 of the models we track fully in memory and the NVIDIA GeForce RTX 5070 Ti 22; only the NVIDIA RTX A6000 holds Qwen 3.8 27B. On the models both run fully in memory, the NVIDIA GeForce RTX 5070 Ti is about 1.2× as fast as the NVIDIA RTX A6000 (estimated, median across models).

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

NVIDIA GeForce RTX 5070 Ti

$995

eBay

Add to build
NVIDIA RTX A6000

Not priced here

Add to build

Specifications

SpecNVIDIA GeForce RTX 5070 TiNVIDIA RTX A6000
GPUGeForce RTX 5070 TiRTX A6000
Memory16 GB GDDR748 GB GDDR6
Memory bus256-bit384-bit
Memory bandwidth896 GB/s768 GB/s
FP16 compute88 TFLOPS38.71 TFLOPS
SoftwareCUDACUDA
Board power300 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 5070 TiNVIDIA RTX A6000
Qwen 3.8 27Boffload13tok/sQ8_0
Qwen 3.5 122B-A10Bwon't run43tok/sUD-Q2_K_XL
Qwen 3.6 27Boffload13tok/sQ8_0
Gemma 4 31Boffload11tok/sQ8_0
Mistral Small 4 119Bwon't run53tok/sUD-Q2_K_XL
Qwen 3.6 35B-A3Boffload49tok/sQ8_0
Qwen 3.5 27Boffload13tok/sQ8_0
Gemma 4 26B-A4Boffload41tok/sQ8_0
Qwen 3.5 35B-A3Boffload49tok/sQ8_0
GLM 4.7 Flashoffload23tok/sQ8_0
Devstral Small 2 24Boffload10tok/sQ8_0
Nemotron 3.5 Lightning 30B-A3Boffload68tok/sUD-Q6_K_XL
gpt-oss 20B60tok/sQ4_K_M51tok/sMXFP4
Kimi Linear 48B-A3Boffload80tok/sQ6_K
Qwen3 Coder 30B-A3Boffload18tok/sQ8_0
Gemma 4 12B45tok/sQ6_K31tok/sQ8_0
Qwen 3.5 9B40tok/sQ8_034tok/sQ8_0
Gemma 3 27Boffload12tok/sQ8_0
Mistral Small 3.2 24Boffload10tok/sQ8_0
Ministral 3 14Boffload13tok/sQ8_0
Qwen3 14Boffload13tok/sQ8_0
DeepSeek R1 0528 Qwen3 8Boffload17tok/sQ8_0
Qwen3 8Boffload17tok/sQ8_0
Phi-4 14Boffload11tok/sQ8_0
Qwen 3.5 4B63tok/sQ8_054tok/sQ8_0
Ministral 3 8Boffload18tok/sQ8_0
Granite 4.2 8Boffload15tok/sQ8_0
Gemma 4 E4B46tok/sQ8_040tok/sQ8_0
LFM2.5 8B-A1B46tok/sUD-Q6_K38tok/sQ8_0
Llama 3.1 8Boffload20tok/sQ6_K
Qwen2.5 0.5B259tok/sQ8_0222tok/sQ8_0
Qwen3.5 0.8B228tok/sQ8_0195tok/sQ8_0
Qwen3.5 2B181tok/sQ5_K_M155tok/sQ5_K_M
DeepSeek R1 Distill Qwen 1.5B98tok/sQ8_084tok/sQ8_0
Mellum2.1 12B A2.5B Thinking73tok/sQ8_062tok/sQ8_0
Spark X2.5 4B70tok/smlx-4bit60tok/smlx-4bit
MiniCPM5 2B67tok/sQ8_057tok/sQ8_0
FrogNano 4B 260962tok/sQ8_053tok/sQ8_0
Kimi VL A3B Thinking 250656tok/sQ4_K_M40tok/sQ8_0
LFM2.5 2.6B51tok/sQ8_044tok/sQ8_0
LFM2.5 VL 3B51tok/sQ8_044tok/sQ8_0
GLM 4.6V Flash48tok/sQ4_K_M31tok/sQ8_0
GLM 4 9B 041441tok/smlx-5bit35tok/smlx-5bit
DeepSeek R1 Distill Qwen 7B40tok/sQ6_K31tok/sQ8_0
granite vision 4.1 4b39tok/sQ8_033tok/sQ8_0
granite 4.2 3b38tok/sQ8_033tok/sQ8_0
Kimi VL A3Boffload39tok/smlx-8bit
Ministral 3 3B Instruct 2512offload27tok/sQ8_0
GLM 4.1V 9B Thinkingoffload28tok/smlx-8bit
Muse Glimmer 30Boffload14tok/sQ8_0
DeepSeek R1 Distill Llama 8Boffload19tok/sQ8_0
LFM2.5 24B A2Boffload22tok/sQ8_0
Ministral 3 8B Reasoning 2512offload18tok/sQ8_0
Qwen3.8 27B OBLITERATEDoffload16tok/smlx-4bit
Ministral 3 14B Reasoning 2512offload13tok/sQ8_0
DeepSeek R1 Distill Qwen 14Boffload12tok/sQ8_0
Devstral Small 2507offload10tok/sQ8_0
Magistral Small 2506offload10tok/sQ8_0
Magistral Small 2507offload10tok/sQ8_0
Magistral Small 2509offload10tok/sQ8_0
Nemotron 3 Super 120B-A12Bwon't runoffload
Qwen3 Coder Nextwon't runoffload
gpt-oss 120Bwon't runoffload
GLM 4.5 Airwon't runoffload
Qwen3 Next 80B-A3Bwon't runoffload
Qwen3 32Bwon't runoffload
DeepSeek R1 Distill Qwen 32Bwon't runoffload
granite 4.2 30bwon'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
Llama 3.3 70Bwon't runwon't run
DeepSeek R1won't runwon't run
DeepSeek R1 0528won't runwon't run
DeepSeek R1 Distill Llama 70Bwon'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 4.5Vwon't runwon't run
GLM 4.6won't runwon't run
GLM 4.6Vwon'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 Dev 72Bwon'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 5070 Ti or the NVIDIA RTX A6000?
Choose the NVIDIA RTX A6000 to run bigger models. Choose the NVIDIA GeForce RTX 5070 Ti for speed.
Which can run bigger models, the NVIDIA GeForce RTX 5070 Ti or the NVIDIA RTX A6000?
At 128k context the NVIDIA RTX A6000 runs 60 of the models we track fully in memory and the NVIDIA GeForce RTX 5070 Ti 22; only the NVIDIA RTX A6000 holds Qwen 3.8 27B.
Which is faster for local AI, the NVIDIA GeForce RTX 5070 Ti or the NVIDIA RTX A6000?
On the models both run fully in memory, the NVIDIA GeForce RTX 5070 Ti is about 1.2× as fast as the NVIDIA RTX A6000 (estimated, median across models).

More comparisons

More hardware for local AI