TEST #3069 RESULTS
03/13/2026 - 5:48 AM
ACCELERATOR
MODEL
43.3
tokens/s
generation
800
ms
time to first token
1695
tokens/s
prompt
451
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core Ultra 9 185H (alderlake)
RAM
62.3GB
OS
Linux
Kernel Release
6.18.7-76061807-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #202601231045~1769703228~24.04~cb87b5b SMP PREEMPT_DYNAMIC Thu J
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1979
tokens/s
44.9
tokens/s
540
ms
pp4096+tg256
1495
tokens/s
39.5
tokens/s
2.77
sec
pp2048+tg256
1779
tokens/s
43.0
tokens/s
1.17
sec
pp2048+tg768
1785
tokens/s
42.5
tokens/s
1.17
sec
pp1024+tg1024
1987
tokens/s
44.1
tokens/s
538
ms
pp1280+tg3072
1886
tokens/s
41.5
tokens/s
701
ms
pp384+tg1152
2168
tokens/s
44.7
tokens/s
199
ms
pp64+tg1024
1699
tokens/s
45.3
tokens/s
59
ms
pp16+tg1536
475
tokens/s
44.6
tokens/s
55
ms