Machine

Apple M5 Max · Apple M5 Max #9a752a52

12 runs of 11 models since Aug 24, 2026. Last run Aug 24, 2026.

ShareRedditX
12
runs
11
models
11
configs
1
configs run 2+ times

Hardware

GPU
Apple M5 Max 128 GB
CPU
Apple M5 Max 18 cores
RAM
128 GB
OS
macOS 26.5.1
Tools
oMLX 11 · Ollama 1
Client
0.4.43
All Apple M5 Max benchmarks

Models on this machine

One dot per model, at its best config. Up is better answers, right is faster. Time to finish counts every token the model wrote, thinking included, so a model that talks a lot ends up further left.

Best trade-off: quality 60+, and nothing else here is both faster and betterOther models · dot size = number of runs
below 20 tok/s: slower than you read4050607080901002050100tok/s (log scale) →quality →

Medians of each model's best config, 11 models.

Best config per model

sort

Runs with the same model, quant, tool, context, thinking and KV cache count as one config. Numbers are the median of its runs, and each dot is one run on the 0–100 quality scale. ★ marks the best config with at least 2 rated runs.

Qwen3.8-27B-AWQ-4.85bpw
1 run · 1 config
★ bestoMLX256k ctxthinking default
85.3
1 rated run
31.8 tok/s
Tiel-Coder-35B-A3B-MLX-oQ4e
1 run · 1 config
★ bestoMLX256k ctxthinking default
84.8
1 rated run
65.9 tok/s
Ornith-1.5-35B-A3B-BigBang-oQ8e-mtp
2 runs · 1 config
★ bestoMLX256k ctxthinking default
×2 runs
84.581.5–87.4
steady
103 tok/s
98.6–107
Ornith-1.5-35B-A3B-oQ8e-mtp
1 run · 1 config
★ bestoMLX256k ctxthinking default
84.5
1 rated run
92.1 tok/s
qwen3.8:27b-mlx
1 run · 1 config
★ bestnvfp4Ollama125k ctxthinking default
83.8
1 rated run
40.3 tok/s
Qwen3.8-27B-Cold-Fusion-GAIN-V1.1-oQ4e-fp16-mtp
1 run · 1 config
★ bestFP16oMLX256k ctxthinking default
83.1
1 rated run
27.8 tok/s
Ornith-1.5-35B-A3B-BigBang-oQ4e-mtp
1 run · 1 config
★ bestoMLX256k ctxthinking default
79.1
1 rated run
125 tok/s
Laguna-S-2.1-NVFP4-mlx
1 run · 1 config
★ bestoMLX256k ctxthinking on
76.7
1 rated run
36.6 tok/s
GLM-4.5-Iceblink-v3-106B-A12B-mlx-mxfp4
1 run · 1 config
★ bestoMLX128k ctxthinking default
74.2
1 rated run
41.1 tok/s
Artemis-31B-v1.1-oQ4e
1 run · 1 config
★ bestoMLX256k ctxthinking default
68.7
1 rated run
18.7 tok/s

Compare with another machine

Run history