TEST #3339 RESULTS
04/17/2026 - 2:25 PM
ACCELERATOR
MODEL
35.0
tokens/s
generation
1.36
sec
time to first token
965
tokens/s
prompt
291
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
13th Gen Intel Core i9-13900K (alderlake)
RAM
63.7GB
OS
Windows
Kernel Release
10.0
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64
RUNTIME
Name
llamafile
Version
0.9.3
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1136
tokens/s
36.1
tokens/s
929
ms
pp4096+tg256
896
tokens/s
31.3
tokens/s
4.60
sec
pp2048+tg256
1031
tokens/s
34.5
tokens/s
2.02
sec
pp2048+tg768
1027
tokens/s
33.9
tokens/s
2.02
sec
pp1024+tg1024
1114
tokens/s
35.4
tokens/s
947
ms
pp1280+tg3072
1078
tokens/s
33.3
tokens/s
1.21
sec
pp384+tg1152
1206
tokens/s
36.4
tokens/s
345
ms
pp64+tg1024
982
tokens/s
37.0
tokens/s
94
ms
pp16+tg1536
214
tokens/s
36.7
tokens/s
101
ms