RTX 3090 vs RTX 4070

For running local LLMs · 5 reports across 3 models

Side A

Vendor: nvidia
VRAM: 24GB
Memory: Discrete

Side B

Vendor: nvidia
VRAM: 12GB
Memory: Discrete

Tokens per second by model

Model	RTX 3090	RTX 4070
Gemma 4up to 31B	—	55.0n=2
Qwen3.6up to 35B	66.0n=2	—
Qwen2.5	28.0n=1	—

More comparisons