can-i-run-this-llm

GPU comparison

NVIDIA RTX 3060 (12GB) vs NVIDIA RTX 4070 (12GB) for local LLMs

How the two stack up for running open-source LLMs locally — memory, bandwidth, price, and how many of the 48 tracked models each one can run.

NVIDIA RTX 3060 (12GB)NVIDIA RTX 4070 (12GB)
Memory12.0 GB12.0 GB
Bandwidth360 GB/s504 GB/s
Price (approx)$329$549
LLMs it runs19 of 4819 of 48
Best model it runsDevstral Small 2 24B · 26–39 tok/sDevstral Small 2 24B · 35–53 tok/s

The NVIDIA RTX 3060 (12GB) and NVIDIA RTX 4070 (12GB) run the same set of tracked models; pick on price and ecosystem. The NVIDIA RTX 4070 (12GB) has more memory bandwidth (504 GB/s), so it generates tokens faster at the same model and quant.

What the NVIDIA RTX 3060 (12GB) runs →What the NVIDIA RTX 4070 (12GB) runs →Open the calculator →