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 ↓
65811 M4 (10c) 16 GB Qwen3-VL-Embedding-2B 4bit 4k 744.7 59.5 26-03-10
65812 M4 (10c) 16 GB Qwen3-VL-Embedding-2B 4bit 1k 818.5 63.9 26-03-10
65813 M1 Max (32c) 64 GB Qwen3.5-35B-A3B 8bit 4k 572.8 46.8 26-03-10
65814 M1 Max (32c) 64 GB Qwen3.5-35B-A3B 8bit 1k 523.0 50.0 26-03-10
65815 M4 (10c) 24 GB Qwen3.5-4B 4bit 4k 406.9 37.5 26-03-10
65816 M4 (10c) 24 GB Qwen3.5-4B 4bit 1k 412.1 39.5 26-03-10
65817 M4 (10c) 16 GB Qwen3-VL-Embedding-2B 4bit 4k 734.8 50.0 26-03-10
65818 M4 (10c) 16 GB Qwen3-VL-Embedding-2B 4bit 1k 888.6 62.8 26-03-10
65819 M1 Max (32c) 64 GB Qwen3.5-27B 4bit 4k 88.9 18.2 26-03-10
65820 M1 Max (32c) 64 GB Qwen3.5-27B 4bit 1k 88.4 19.2 26-03-10
68,168 results · Page 6582 of 6817