FitMyLLM

Running LLMs on NVIDIA Tesla M2050: What You Can Run

The NVIDIA Tesla M2050 packs 3GB of GDDR5 with 148 GB/s of memory bandwidth, making it an entry-level 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.

Loading the model tables…

Running LLMs on NVIDIA Tesla M2050: Complete Guide

The NVIDIA Tesla M2050 has 3GB VRAM and 148 GB/s memory bandwidth, making it an entry-level option for running local AI models. This guide covers which LLMs fit, expected tok/s performance, and recommended settings.

See the full NVIDIA Tesla M2050 specs page for detailed specifications, all compatible models, and speed estimates.