Community Benchmarks

Download oMLX →

Real performance numbers from oMLX users around the world. Submit yours from the app (v0.2.6+).

Reset filters
# Chip RAM Model Quant Ctx PP tok/s TG tok/s Date ↓
12971 M1 Max (32c) 32 GB Qwen3.5-9B 8bit 64k 205.4 14.8 26-03-09
12972 M3 Max (40c) 64 GB Qwen3.5-9B 4bit 64k 523.0 35.8 26-03-09
12973 M4 Pro (20c) 64 GB Qwen3.5-35B-A3B 4bit 64k 511.5 33.2 26-03-09
12974 M3 Ultra (60c) 96 GB Qwen3.5-0.8B9bit 8bit 64k 4,818 122.4 26-03-09
12975 M1 Max (32c) 32 GB Qwen3.5-9B 8bit 64k 220.1 15.6 26-03-09
12976 M1 Ultra (48c) 128 GB Qwen3.5-35B-A3B-Heretic 4bit 64k 586.3 29.9 26-03-09
12977 M2 Ultra (60c) 64 GB GLM-4.7-Flash 6bit 64k 195.2 15.8 26-03-09
12978 M2 Ultra (76c) 192 GB Qwen3.5-0.8B 8bit 64k 3,908 84.5 26-03-09
12979 M2 Ultra (76c) 192 GB Qwen3.5-0.8B 4bit 64k 3,882 87.8 26-03-09
12980 M4 (10c) 24 GB Qwen3.5-4B 4bit 64k 247.3 14.5 26-03-09
13,020 results · Page 1298 of 1302