Can a Mac Studio M3 Ultra (96GB) run GPT-OSS 120B (MoE)?

Just about — GPT-OSS 120B (MoE) runs at FP16 (65.4 GB) at about 89–133 tok/s on a Mac Studio M3 Ultra (96GB).

How much memory does GPT-OSS 120B (MoE) need?

GPT-OSS 120B (MoE) needs about 63.5 GB at 4K context at its smallest quantization. A Mac Studio M3 Ultra (96GB) has 96.0 GB.

Tight fit: GPT-OSS 120B (MoE) on a Mac Studio M3 Ultra (96GB) at FP16

Yes, but it's tight.

Tight / slow

Best fit: GPT-OSS 120B (MoE) at FP16 · 65.4 GB weights · 89–133 tok/s on a Mac Studio M3 Ultra (96GB).

Arena Text Overall · style-controlled

1,352FP16 untested

gpt-oss-120b · published configuration reference · FP16 not tested

Arena Text Overall (style-controlled), exact published entry gpt-oss-120b, 2026-10-08: reference rating 1,351.8. Published confidence bounds 1,347.5–1,356.2. 30,218 votes. Published gpt-oss-120b configuration reference. Reasoning mode is not disclosed by the Arena dataset. Tested precision, runtime and sampling settings are not disclosed in this dataset. Selected FP16 has not been measured by this Arena entry. No quantization quality adjustment is applied.

Snapshot published 2026-10-08 · snapshot imported 2026-10-11 UTC · validated last-good data · manual refresh

Arena leaderboard dataset · Arena / lmarena-ai, Arena Leaderboard Dataset (CC BY 4.0). Selected Text Overall style-controlled rows; field names adapted, ratings unchanged. · dataset CC BY 4.0 · revision e16f90232986

Total memory at 4K context: 66.3 GB; Weights + KV cache + runtime overhead
Estimated generation: 89–133 tok/s
Estimated first token: ~2.2s

Every quantization on this machine

On small screens, each quant lists every metric below its name. On wider screens, scroll sideways for total memory, speed and fit. Keyboard: focus the table, then use the arrow keys. Quant names stay visible.

GPT-OSS 120B (MoE) quantizations on Mac Studio M3 Ultra (96GB), estimated at 4K context. Weights and total memory are shown separately.
Quant	Weights	Total memory	Speed	Fit
Q2_K	62.6 GB	63.5 GB	90–136 tok/s	Tight / slow
Q3_K_M	62.6 GB	63.6 GB	90–136 tok/s	Tight / slow
Q4_K_M	62.8 GB	63.7 GB	90–136 tok/s	Tight / slow
Q5_K_M	62.9 GB	63.8 GB	90–135 tok/s	Tight / slow
Q6_K	63.3 GB	64.2 GB	90–135 tok/s	Tight / slow
Q8_0	63.4 GB	64.3 GB	90–135 tok/s	Tight / slow
FP16	65.4 GB	66.3 GB	89–133 tok/s	Tight / slow

Open the full calculator →Everything a Mac Studio M3 Ultra (96GB) runs GPT-OSS 120B (MoE) on other machines What to buy to run it →

Ran GPT-OSS 120B (MoE) on a Mac Studio M3 Ultra (96GB)? Add your real numbers →