TEST #397 RESULTS
04/06/2025 - 5:17 PM
ACCELERATOR
MODEL
48.7
tokens/s
generation
2.16
sec
time to first token
534
tokens/s
prompt
229
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M1 Ultra 16P+4E
RAM
128GB
OS
Darwin
Kernel Release
24.3.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 24.3.0: Thu Jan 2 20:24:16 PST 2025; root:xnu-11215.81.4~3/RELEASE_ARM64_T6000
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
645
tokens/s
52.1
tokens/s
1.61
sec
pp4096+tg256
618
tokens/s
39.3
tokens/s
6.65
sec
pp2048+tg256
641
tokens/s
46.8
tokens/s
3.22
sec
pp2048+tg768
641
tokens/s
45.7
tokens/s
3.22
sec
pp1024+tg1024
637
tokens/s
49.7
tokens/s
1.63
sec
pp1280+tg3072
640
tokens/s
44.0
tokens/s
2.02
sec
pp384+tg1152
570
tokens/s
52.7
tokens/s
692
ms
pp64+tg1024
320
tokens/s
54.6
tokens/s
217
ms
pp16+tg1536
90
tokens/s
53.6
tokens/s
196
ms