Model fit guide / MI300X 192GBBenchmark-first pickUpdated June 3, 2026
192 GB VRAM - AMD datacenter 192GB

Best local AI model for MI300X 192GB.

A memory-rich single-GPU target for large current MoE, a better row than "70B+." Use 192GB where it avoids tensor-parallel complexity or improves batch and context economics. The constraint is ROCm/runtime maturity, not raw VRAM.

01 / Recommendation

Run this size class.

Recommended default

GLM-5 / Kimi K2.6 / MiniMax-M2-class

Use FP8/INT8 where supported, or runtime-specific quantization. This is the highest-scoring current open-weight model that fits this card cleanly, selected by benchmark then fit then freshness, not by parameter count.

Benchmark anchor

MiniMax-M2 claims #1 open-source global composite by Artificial Analysis at release; verify against your own target benchmark.

Evidence

MiniMax-M2 release claims a #1 open-source composite; treat as a candidate and verify on your target benchmark. MI300X is positioned around memory capacity and runtime support.

02 / Alternates

Other realistic picks.

GLM-5

Kimi K2.6

DeepSeek V4-class large MoE

03 / More GPUs

Compare another card.

RTX 3060 12GBRTX 4060 Ti 16GBRTX 5080 16GBRTX 3090 24GBRTX 4090 24GB