TEST #2442 RESULTS
02/01/2026 - 5:44 AM
ACCELERATOR
MODEL
47.4
tokens/s
generation
2.22
sec
time to first token
553
tokens/s
prompt
228
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M4 Max 10P+4E
RAM
36GB
OS
Darwin
Kernel Release
25.3.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 25.3.0: Thu Jan 22 21:54:32 PST 2026; root:xnu-12377.80.0.501.1~2/RELEASE_ARM64_T6041
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
631
tokens/s
52.1
tokens/s
1.64
sec
pp4096+tg256
594
tokens/s
37.8
tokens/s
6.93
sec
pp2048+tg256
625
tokens/s
45.6
tokens/s
3.30
sec
pp2048+tg768
625
tokens/s
44.8
tokens/s
3.30
sec
pp1024+tg1024
609
tokens/s
49.2
tokens/s
1.70
sec
pp1280+tg3072
576
tokens/s
43.0
tokens/s
2.24
sec
pp384+tg1152
643
tokens/s
52.6
tokens/s
614
ms
pp64+tg1024
504
tokens/s
52.3
tokens/s
144
ms
pp16+tg1536
171
tokens/s
49.6
tokens/s
112
ms