Skip to content
can-i-run-this-llm

Hardware comparison

AMD Ryzen AI Max+ 395 “Strix Halo” (64GB) vs AMD Ryzen AI Max+ 395 “Strix Halo” (128GB) for local LLMs

How the two stack up for running open-source LLMs locally — memory, bandwidth, price, and how many of the 69 tracked models each one can run.

Scroll sideways to compare both machines. Keyboard: focus the table, then use the arrow keys. Metric names stay visible.

AMD Ryzen AI Max+ 395 “Strix Halo” (64GB) compared with AMD Ryzen AI Max+ 395 “Strix Halo” (128GB) for local LLM inference
HardwareAMD Ryzen AI Max+ 395 “Strix Halo” (64GB)AMD Ryzen AI Max+ 395 “Strix Halo” (128GB)
Memory64.0 GB128.0 GB (better for this comparison)
Bandwidth256 GB/s256 GB/s
Price (approx)$1,999 (better for this comparison)Complete system$2,999Complete system
LLMs it runs57 of 6962 of 69 (better for this comparison)
Best model it runsGemma 4 31BQ2_K · estimated 12–19 tok/sGemma 4 31BQ2_K · estimated 12–19 tok/s

Both prices cover complete systems. Prices are approximate catalog estimates; check current configurations and prices before buying.

Both estimated speeds use the same model and quant shown above. Estimates use 4,096 context tokens, llama.cpp and one stream.

The AMD Ryzen AI Max+ 395 “Strix Halo” (128GB) runs 5 more of the tracked models (62 vs 57), thanks to its 128.0 GB of memory.

What the AMD Ryzen AI Max+ 395 “Strix Halo” (64GB) runs →What the AMD Ryzen AI Max+ 395 “Strix Halo” (128GB) runs →Open the calculator →