Community Benchmarks

Download oMLX →

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

# Chip RAM Model Quant Ctx PP tok/s TG tok/s Date ↓
6801 M2 Ultra (60c) 64 GB Qwen3-Next-80B-A3B-Instru... 4bit 16k 822.4 49.4 26-03-09
6802 M2 Ultra (60c) 64 GB Qwen3-Next-80B-A3B-Instru... 4bit 8k 852.2 55.4 26-03-09
6803 M2 Ultra (60c) 64 GB Qwen3-Next-80B-A3B-Instru... 4bit 4k 848.4 59.5 26-03-09
6804 M2 Ultra (60c) 64 GB Qwen3-Next-80B-A3B-Instru... 4bit 1k 729.9 61.8 26-03-09
6805 M4 (10c) 16 GB Qwen3.5-4B 4bit 1k 405.8 36.7 26-03-09
6806 M4 (10c) 16 GB Qwen3.5-4B 4bit 16k 351.7 27.9 26-03-09
6807 M4 (10c) 16 GB Qwen3.5-4B 4bit 8k 390.9 33.4 26-03-09
6808 M4 (10c) 16 GB Qwen3.5-4B 4bit 4k 403.9 34.7 26-03-09
6809 M4 (10c) 16 GB Qwen3.5-4B 4bit 1k 406.1 37.5 26-03-09
6810 M4 Pro (16c) 64 GB Qwen3.5-0.8B 8bit 4k 3,453 201.8 26-03-09
7,296 results · Page 681 of 730