readygpu

NVIDIA GeForce RTX 4070 vs NVIDIA GeForce RTX 4090

For running models locally, priced in the US. At 128k context the first runs 13 models fully in memory, the second 31.

In short

The NVIDIA GeForce RTX 4090 has more memory for models: 24 GB against 12 GB. At 128k context the NVIDIA GeForce RTX 4090 runs 31 of the models we track fully in memory and the NVIDIA GeForce RTX 4070 13; only the NVIDIA GeForce RTX 4090 holds Gemma 4 26B-A4B. On the models both run fully in memory, the NVIDIA GeForce RTX 4090 is about 2.0× as fast as the NVIDIA GeForce RTX 4070 (estimated, median across models).

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

NVIDIA GeForce RTX 4070

Not priced here

Add to build
NVIDIA GeForce RTX 4090

$2,073

eBay

Add to build

Specifications

SpecNVIDIA GeForce RTX 4070NVIDIA GeForce RTX 4090
GPUGeForce RTX 4070GeForce RTX 4090
Memory12 GB GDDR6X24 GB GDDR6X
Memory bus192-bit384-bit
Memory bandwidth504.2 GB/s1,008 GB/s
FP16 compute29.146 TFLOPS165 TFLOPS
SoftwareCUDACUDA
Board power200 W450 W
PCIe generationGen 4Gen 4
PCIe lanesx16x16

What each runs

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

ModelNVIDIA GeForce RTX 4070NVIDIA GeForce RTX 4090
Gemma 4 26B-A4Boffload64tok/sQ5_K_M
gpt-oss 20Boffload66tok/sMXFP4
Gemma 4 12B33tok/sQ4_K_M41tok/sQ8_0
Qwen 3.5 9B31tok/sQ4_K_M45tok/sQ8_0
Qwen 3.5 4B35tok/sQ8_071tok/sQ8_0
Ministral 3 8Boffload27tok/sQ4_K_M
Gemma 4 E4B34tok/sQ5_K_M52tok/sQ8_0
LFM2.5 8B-A1Boffload50tok/sQ8_0
Llama 3.1 8Boffload27tok/sQ5_K_M
Qwen2.5 0.5B146tok/sQ8_0292tok/sQ8_0
Qwen3.5 0.8B128tok/sQ8_0256tok/sQ8_0
Qwen3.5 2B102tok/sQ5_K_M204tok/sQ5_K_M
DeepSeek R1 Distill Qwen 1.5B55tok/sQ8_0111tok/sQ8_0
Spark X2.5 4B39tok/smlx-4bit79tok/smlx-4bit
MiniCPM5 2B38tok/sQ8_075tok/sQ8_0
FrogNano 4B 260935tok/sQ8_070tok/sQ8_0
LFM2.5 2.6B30tok/sQ6_K57tok/sQ8_0
LFM2.5 VL 3B30tok/sQ6_K57tok/sQ8_0
Mellum2.1 12B A2.5B Thinkingoffload82tok/sQ8_0
GLM 4.6V Flashoffload41tok/sQ8_0
DeepSeek R1 Distill Qwen 7Boffload40tok/sQ8_0
GLM 4 9B 0414offload46tok/smlx-5bit
Kimi VL A3B Thinking 2506offload52tok/sQ8_0
granite vision 4.1 4boffload44tok/sQ8_0
granite 4.2 3boffload43tok/sQ8_0
Kimi VL A3Boffload51tok/smlx-8bit
Ministral 3 3B Instruct 2512offload36tok/sQ8_0
GLM 4.1V 9B Thinkingoffload37tok/smlx-8bit
Muse Glimmer 30Boffload29tok/sQ5_K_M
DeepSeek R1 Distill Llama 8Boffload27tok/sQ5_K_M
Ministral 3 8B Reasoning 2512offload27tok/sQ4_K_M
Qwen 3.8 27Boffloadoffload
Qwen 3.5 122B-A10Bwon't runoffload
Qwen 3.6 27Boffloadoffload
Gemma 4 31Boffloadoffload
Mistral Small 4 119Bwon't runoffload
Qwen 3.6 35B-A3Boffloadoffload
Qwen 3.5 27Boffloadoffload
Qwen 3.5 35B-A3Boffloadoffload
GLM 4.7 Flashoffloadoffload
Devstral Small 2 24Boffloadoffload
Nemotron 3.5 Lightning 30B-A3Boffloadoffload
Kimi Linear 48B-A3Boffloadoffload
Qwen3 Coder 30B-A3Boffloadoffload
Gemma 3 27Boffloadoffload
Mistral Small 3.2 24Boffloadoffload
Ministral 3 14Boffloadoffload
Qwen3 14Boffloadoffload
DeepSeek R1 0528 Qwen3 8Boffloadoffload
Qwen3 8Boffloadoffload
Phi-4 14Boffloadoffload
Granite 4.2 8Boffloadoffload
LFM2.5 24B A2Boffloadoffload
Ministral 3 14B Reasoning 2512offloadoffload
Qwen3.8 27B OBLITERATEDoffloadoffload
DeepSeek R1 Distill Qwen 14Boffloadoffload
Devstral Small 2507offloadoffload
Magistral Small 2506offloadoffload
Magistral Small 2507offloadoffload
Magistral Small 2509offloadoffload
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
Nemotron 3 Super 120B-A12Bwon't runwon't run
Qwen3 Coder Nextwon't runwon't run
gpt-oss 120Bwon't runwon't run
GLM 4.5 Airwon't runwon't run
Qwen3 Next 80B-A3Bwon't runwon't run
Qwen3 32Bwon't runwon't run
DeepSeek R1 Distill Qwen 32Bwon'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
granite 4.2 30bwon't runwon't run

Questions

Which is better for local AI, the NVIDIA GeForce RTX 4070 or the NVIDIA GeForce RTX 4090?
Choose the NVIDIA GeForce RTX 4090 to run bigger models. Choose the NVIDIA GeForce RTX 4090 for speed.
Which can run bigger models, the NVIDIA GeForce RTX 4070 or the NVIDIA GeForce RTX 4090?
At 128k context the NVIDIA GeForce RTX 4090 runs 31 of the models we track fully in memory and the NVIDIA GeForce RTX 4070 13; only the NVIDIA GeForce RTX 4090 holds Gemma 4 26B-A4B.
Which is faster for local AI, the NVIDIA GeForce RTX 4070 or the NVIDIA GeForce RTX 4090?
On the models both run fully in memory, the NVIDIA GeForce RTX 4090 is about 2.0× as fast as the NVIDIA GeForce RTX 4070 (estimated, median across models).

More comparisons

More hardware for local AI