Can I run this LLM?
Can a Mac Studio M3 Ultra (512GB) run Gemma 4 26B-A4B (MoE)?
Gemma · Apache 2.0 · estimated for a 512.0 GB machine at 4K context.
Yes — it runs great.
Best fit: Gemma 4 26B-A4B (MoE) at FP16 · 50.5 GB · 48–72 tok/s on a Mac Studio M3 Ultra (512GB).
Every quantization on this machine
| Quant | Size | Speed | Fit |
|---|---|---|---|
| Q3_K_M | 12.7 GB | 100–150 tok/s | Runs great |
| Q4_K_M | 17.0 GB | 89–134 tok/s | Runs great |
| Q5_K_M | 21.2 GB | 80–121 tok/s | Runs great |
| Q6_K | 23.2 GB | 77–115 tok/s | Runs great |
| Q8_0 | 26.9 GB | 71–107 tok/s | Runs great |
| FP16 | 50.5 GB | 48–72 tok/s | Runs great |
Open the full calculator →Everything a Mac Studio M3 Ultra (512GB) runsGemma 4 26B-A4B (MoE) on other machinesWhat to buy to run it →
Ran Gemma 4 26B-A4B (MoE) on a Mac Studio M3 Ultra (512GB)? Add your real numbers →