Qwen
Qwen
Qwen/Qwen3.6-27B

Runs

View all benchmark runs for this model family.

	Quant						Actions

	Quant		Actions

	Quant						Actions
M5 Max	8bit	mlx_lm0.31.3	14.8 tok/s	509.7 tok/s	36.00 GB 28%
M1 Max	q4_k_m	llama.cppb8180	8.3 tok/s	117.8 tok/s	4.06 GB 13%
M1 Max	8bit	mlx_lm0.31.3	11.3 tok/s	90.4 tok/s	31.00 GB 48%
M4 Pro	4bit	mlx_lm0.31.3	15.1 tok/s	100.4 tok/s	18.00 GB 75%
M4 Pro	8bit	mlx_lm0.31.3	8.0 tok/s	113.3 tok/s	33.00 GB 52%
M3 Max	8bit	mlx_lm0.31.2	8.5 tok/s	142.2 tok/s	36.00 GB 28%
M5 Pro	q4_k_m	llama.cppb8920	15.0 tok/s	264.8 tok/s	10.25 GB 16%
M5 Pro	nvfp4	mlx_lm0.31.3	17.7 tok/s	477.8 tok/s	18.00 GB 28%
M5 Max	q4_k_m	llama.cppb8680	25.6 tok/s	477.4 tok/s	9.16 GB 14%

	Quant		Actions
M5 Max	8bit	14.8 tok/s
M1 Max	q4_k_m	8.3 tok/s
M1 Max	8bit	11.3 tok/s
M4 Pro	4bit	15.1 tok/s
M4 Pro	8bit	8.0 tok/s
M3 Max	8bit	8.5 tok/s
M5 Pro	q4_k_m	15.0 tok/s
M5 Pro	nvfp4	17.7 tok/s
M5 Max	q4_k_m	25.6 tok/s