Runs | whatcani.run

Runs

Most recent benchmark runs across all devices and models.

								Actions
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source

			Actions
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source
Source

								Actions
model.gguf	EPYCCPU	llama.cppb8240	300.5 tok/s	1,300.0 tok/s	0.00 GB 0%	Rejected
wcir-model	EPYCCPU	mlx_lm0.30.1	200.0 tok/s	1,000.0 tok/s	N/A	Rejected
Qwen3.8-27B Source	M1 Max	llama.cpp0.3.0-dev	8.1 tok/s	93.7 tok/s	15.50 GB 24%	Verified
Qwen3.8-27B Source	M3 Max	llama.cpp0.3.0-dev	12.0 tok/s	172.3 tok/s	16.54 GB 13%	Verified
Qwen3.8-27B Source	M3 Max	llama.cpp0.3.0-dev	12.6 tok/s	170.9 tok/s	16.64 GB 13%	Verified
Qwen3.5-0.8B Source	M2	mlx_lm0.31.3	115.1 tok/s	1,098.3 tok/s	1.99 GB 12%	Verified
gpt-oss-20b Source	M4 Pro	llama.cpp0.3.0	64.9 tok/s	635.2 tok/s	6.21 GB 26%	Verified
Qwen3.5-9B Source	M4 Pro	llama.cpp0.3.0	29.3 tok/s	332.1 tok/s	4.91 GB 20%	Verified
Gemma 4 E4B IT Source	M1	mlx_lm0.31.3	13.1 tok/s	180.3 tok/s	7.50 GB 47%	Verified
Qwen3.5-0.8B Source	M1	mlx_lm0.31.3	46.1 tok/s	814.3 tok/s	2.46 GB 15%	Verified

			Actions
model.gguf	300.5 tok/s	Rejected
wcir-model	200.0 tok/s	Rejected
Qwen3.8-27B Source	8.1 tok/s	Verified
Qwen3.8-27B Source	12.0 tok/s	Verified
Qwen3.8-27B Source	12.6 tok/s	Verified
Qwen3.5-0.8B Source	115.1 tok/s	Verified
gpt-oss-20b Source	64.9 tok/s	Verified
Qwen3.5-9B Source	29.3 tok/s	Verified
Gemma 4 E4B IT Source	13.1 tok/s	Verified
Qwen3.5-0.8B Source	46.1 tok/s	Verified