Highest Arena reference that fits

Hunyuan Hy3 (MoE)Runs great

Q3_K_M · 137.6 GB weights · 139.7 GB total · 180–271 tok/s · Arena reference 1,454

Arena is a published configuration reference. Selected Q3_K_M is untested by this reference.

Arena reference details

Arena Text Overall · style-controlled

1,454Q3_K_M untested

hy3 · published configuration reference · Q3_K_M not tested

Arena Text Overall (style-controlled), exact published entry hy3, 2026-10-08: reference rating 1,454.2. Published confidence bounds 1,447.9–1,460.5. 11,284 votes. Published hy3 configuration reference. Reasoning mode is not disclosed by the Arena dataset. Tested precision, runtime and sampling settings are not disclosed in this dataset. Selected Q3_K_M 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 300–450 tok/s

Also runs

Open the full calculator →Check another machine

Have a AMD Instinct MI300X (192GB)? Add your real benchmark → to sharpen these numbers for everyone.

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