readygpu

NVIDIA GeForce RTX 3080 10GB vs NVIDIA GeForce RTX 4070

For running models locally, priced in Germany. At 64k context the first runs 18 models fully in memory, the second 21.

In short

The NVIDIA GeForce RTX 4070 has more memory for models: 12 GB against 10 GB. At 64k context the NVIDIA GeForce RTX 4070 runs 21 of the models we track fully in memory and the NVIDIA GeForce RTX 3080 10GB 18; only the NVIDIA GeForce RTX 4070 holds Mellum2.1 12B A2.5B Thinking. On the models both run fully in memory, the NVIDIA GeForce RTX 3080 10GB is about 1.5× as fast as the NVIDIA GeForce RTX 4070 (estimated, median across models). In Germany the NVIDIA GeForce RTX 3080 10GB costs 699 € at eBay and the NVIDIA GeForce RTX 4070 700 € at eBay: the NVIDIA GeForce RTX 3080 10GB is 1 € cheaper.

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

NVIDIA GeForce RTX 3080 10GB

699 €

eBay

Add to build
NVIDIA GeForce RTX 4070

700 €

eBay

Add to build

Specifications

SpecNVIDIA GeForce RTX 3080 10GBNVIDIA GeForce RTX 4070
GPUGeForce RTX 3080 10GBGeForce RTX 4070
Memory10 GB GDDR6X12 GB GDDR6X
Memory bus320-bit192-bit
Memory bandwidth760 GB/s504.2 GB/s
FP16 compute59.5 TFLOPS29.146 TFLOPS
SoftwareCUDACUDA
Board power320 W200 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 3080 10GBNVIDIA GeForce RTX 4070
Gemma 4 12B55tok/sQ4_K_M32tok/sQ5_K_M
Qwen 3.5 9B54tok/sQ5_K_M32tok/sQ6_K
Qwen 3.5 4B71tok/sQ8_047tok/sQ8_0
Gemma 4 E4B64tok/sQ5_K_M31tok/sQ8_0
LFM2.5 8B-A1B73tok/sMXFP446tok/sUD-Q6_K
Qwen2.5 0.5B353tok/sQ8_0234tok/sQ8_0
Qwen3.5 0.8B288tok/sQ8_0191tok/sQ8_0
Qwen3.5 2B209tok/sQ5_K_M138tok/sQ5_K_M
DeepSeek R1 Distill Qwen 1.5B125tok/sQ8_083tok/sQ8_0
MiniCPM5 2B86tok/sQ8_057tok/sQ8_0
Spark X2.5 4B86tok/smlx-4bit57tok/smlx-4bit
FrogNano 4B 260969tok/sQ8_046tok/sQ8_0
LFM2.5 2.6B68tok/sQ8_045tok/sQ8_0
LFM2.5 VL 3B68tok/sQ8_045tok/sQ8_0
granite vision 4.1 4b60tok/sQ5_K_M35tok/sQ8_0
DeepSeek R1 Distill Qwen 7B56tok/sQ4_K_M31tok/sQ6_K
granite 4.2 3b56tok/sQ6_K34tok/sQ8_0
GLM 4.6V Flash53tok/sQ4_K_M35tok/sQ4_K_M
Mellum2.1 12B A2.5B Thinkingoffload62tok/sQ5_K_M
Ministral 3 3B Instruct 2512offload29tok/sQ8_0
GLM 4 9B 0414offload29tok/smlx-5bit
Qwen 3.8 27Boffloadoffload
Qwen 3.6 27Boffloadoffload
Gemma 4 31Boffloadoffload
Qwen 3.6 35B-A3Boffloadoffload
Qwen 3.5 27Boffloadoffload
Gemma 4 26B-A4Boffloadoffload
Qwen 3.5 35B-A3Boffloadoffload
GLM 4.7 Flashoffloadoffload
Devstral Small 2 24Boffloadoffload
Nemotron 3.5 Lightning 30B-A3Boffloadoffload
gpt-oss 20Boffloadoffload
Kimi Linear 48B-A3Boffloadoffload
Qwen3 Coder 30B-A3Boffloadoffload
Qwen3 32Bwon't runoffload
Gemma 3 27Boffloadoffload
Mistral Small 3.2 24Boffloadoffload
DeepSeek R1 Distill Qwen 32Bwon't runoffload
Ministral 3 14Boffloadoffload
Qwen3 14Boffloadoffload
DeepSeek R1 0528 Qwen3 8Boffloadoffload
Qwen3 8Boffloadoffload
Phi-4 14Boffloadoffload
Ministral 3 8Boffloadoffload
Granite 4.2 8Boffloadoffload
Llama 3.1 8Boffloadoffload
Kimi VL A3B Thinking 2506offloadoffload
Kimi VL A3Boffloadoffload
DeepSeek R1 Distill Llama 8Boffloadoffload
Ministral 3 8B Reasoning 2512offloadoffload
GLM 4.1V 9B Thinkingoffloadoffload
LFM2.5 24B A2Boffloadoffload
Muse Glimmer 30Boffloadoffload
Ministral 3 14B Reasoning 2512offloadoffload
DeepSeek R1 Distill Qwen 14Boffloadoffload
Devstral Small 2507offloadoffload
Magistral Small 2506offloadoffload
Magistral Small 2507offloadoffload
Magistral Small 2509offloadoffload
Qwen3.8 27B OBLITERATEDoffloadoffload
granite 4.2 30boffloadoffload
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
Qwen 3.5 122B-A10Bwon'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
Mistral Small 4 119Bwon't runwon't run
GLM 4.5 Airwon't runwon't run
Qwen3 Next 80B-A3Bwon'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 3080 10GB or the NVIDIA GeForce RTX 4070?
Choose the NVIDIA GeForce RTX 4070 to run bigger models. Choose the NVIDIA GeForce RTX 3080 10GB for speed. Choose the NVIDIA GeForce RTX 3080 10GB to spend less.
Which can run bigger models, the NVIDIA GeForce RTX 3080 10GB or the NVIDIA GeForce RTX 4070?
At 64k context the NVIDIA GeForce RTX 4070 runs 21 of the models we track fully in memory and the NVIDIA GeForce RTX 3080 10GB 18; only the NVIDIA GeForce RTX 4070 holds Mellum2.1 12B A2.5B Thinking.
Which is faster for local AI, the NVIDIA GeForce RTX 3080 10GB or the NVIDIA GeForce RTX 4070?
On the models both run fully in memory, the NVIDIA GeForce RTX 3080 10GB is about 1.5× as fast as the NVIDIA GeForce RTX 4070 (estimated, median across models).
What is the price difference between the NVIDIA GeForce RTX 3080 10GB and the NVIDIA GeForce RTX 4070?
In Germany the NVIDIA GeForce RTX 3080 10GB costs 699 € at eBay and the NVIDIA GeForce RTX 4070 700 € at eBay: the NVIDIA GeForce RTX 3080 10GB is 1 € cheaper.

More comparisons

More hardware for local AI