M3 Ultra 96GB vs M4 Pro 48GB
For running local LLMs · 6 reports across 4 models
Highest reported generation speed by model family
These maxima come from different community setups. Model size, quantization, GPU count, offloading and concurrency may differ. They do not establish which GPU is faster for the same workload. Check the comparison conditions →
| Model | M3 Ultra 96GB | M4 Pro 48GB |
|---|---|---|
| Museup to 30B | — | 18.0n=2 |
| Qwen3.8up to 2400B | —n=1 | 20.3n=1 |
| GLM-5.3up to 321B | — | 1.8n=1 |
| LingBot-World-V2up to 1B | — | —n=1 |