The Google Tensor G3 GPU packs 12GB of Shared with 34 GB/s of memory bandwidth, making it a capable option for running local AI models. Here is which models fit, how fast they run, and what quantisation gives the best balance of quality and speed.
Already have it? Find the best model for it. Still choosing? See what a budget buys. Or the full specifications.
The Google Tensor G3 GPU has 12GB VRAM and 34 GB/s memory bandwidth, making it a capable option for running local AI models. This guide covers which LLMs fit, expected tok/s performance, and recommended settings.
See the full Google Tensor G3 GPU specs page for detailed specifications, all compatible models, and speed estimates.