TEST #2084 RESULTS
01/03/2026 - 4:35 PM
ACCELERATOR
MODEL
21.0
tokens/s
generation
1.56
sec
time to first token
944
tokens/s
prompt
234
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core Ultra 7 265H (arrowlake)
RAM
30.3GB
OS
Linux
Kernel Release
6.17.0-8-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #8-Ubuntu SMP PREEMPT_DYNAMIC Fri Nov 14 21:44:46 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
806
tokens/s
10.8
tokens/s
1.38
sec
pp4096+tg256
715
tokens/s
22.5
tokens/s
5.77
sec
pp2048+tg256
1006
tokens/s
22.9
tokens/s
2.07
sec
pp2048+tg768
986
tokens/s
22.0
tokens/s
2.11
sec
pp1024+tg1024
1106
tokens/s
22.6
tokens/s
956
ms
pp1280+tg3072
1073
tokens/s
21.5
tokens/s
1.22
sec
pp384+tg1152
1097
tokens/s
22.7
tokens/s
380
ms
pp64+tg1024
1150
tokens/s
21.8
tokens/s
79
ms
pp16+tg1536
553
tokens/s
22.6
tokens/s
54
ms