TEST #3583 RESULTS
05/21/2026 - 3:06 AM
ACCELERATOR
MODEL
42.2
tokens/s
generation
2.47
sec
time to first token
475
tokens/s
prompt
201
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
25.5.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 25.5.0: Mon Apr 27 20:41:15 PDT 2026; root:xnu-12377.121.6~2/RELEASE_ARM64_T6041
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
658
tokens/s
53.4
tokens/s
1.57
sec
pp4096+tg256
579
tokens/s
37.0
tokens/s
7.10
sec
pp2048+tg256
557
tokens/s
44.7
tokens/s
3.70
sec
pp2048+tg768
522
tokens/s
43.4
tokens/s
3.94
sec
pp1024+tg1024
507
tokens/s
47.5
tokens/s
2.04
sec
pp1280+tg3072
478
tokens/s
41.2
tokens/s
2.70
sec
pp384+tg1152
513
tokens/s
41.7
tokens/s
769
ms
pp64+tg1024
349
tokens/s
36.2
tokens/s
206
ms
pp16+tg1536
115
tokens/s
34.8
tokens/s
164
ms