Highest Arena reference that fits

Gemma 4 31BRuns great

Q6_K · 25.2 GB weights · 27.2 GB total · 14–22 tok/s · Arena reference 1,452

Arena is a published configuration reference. Selected Q6_K is untested by this reference.

Arena reference details

Arena Text Overall · style-controlled

1,452Q6_K untested

gemma-4-31b · published configuration reference · Q6_K not tested

Arena Text Overall (style-controlled), exact published entry gemma-4-31b, 2026-10-08: reference rating 1,452.4. Published confidence bounds 1,444.9–1,459.8. 6,145 votes. Published gemma-4-31b configuration reference. Reasoning mode is not disclosed by the Arena dataset. Tested precision, runtime and sampling settings are not disclosed in this dataset. Selected Q6_K 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

Fastest that fits: LFM2.5 1.2B at 86–129 tok/s

Also runs

Open the full calculator →Check another machine

Compare against

vs MacBook Pro / Mac Studio M4 Max (128GB)vs Mac Studio M5 Ultra (96GB)vs NVIDIA DGX Spark (GB10, 128GB)vs NVIDIA RTX 5090 (32GB)vs NVIDIA RTX 6000 Ada (48GB)vs MacBook Pro / Mac Studio M4 Max (48GB)vs AMD Ryzen AI Max+ 395 “Strix Halo” (128GB)vs NVIDIA RTX 4090 (24GB)all comparisons →

Have a MacBook Pro / Mac Studio M5 Max (128GB)? Add your real benchmark → to sharpen these numbers for everyone.

Estimates from device specs and model architectures, refined by community benchmarks. See the methodology.