TEST #3253 RESULTS
04/07/2026 - 6:20 PM
ACCELERATOR
MODEL
2.8
tokens/s
generation
152.10
sec
time to first token
9
tokens/s
prompt
6
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core i7-8650U CPU @ 1.90GHz (skylake)
RAM
15.9GB
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
12
tokens/s
3.1
tokens/s
88.64
sec
pp4096+tg256
8
tokens/s
2.4
tokens/s
511.58
sec
pp2048+tg256
9
tokens/s
2.7
tokens/s
231.49
sec
pp2048+tg768
9
tokens/s
2.7
tokens/s
239.16
sec
pp1024+tg1024
9
tokens/s
2.9
tokens/s
109.41
sec
pp1280+tg3072
9
tokens/s
2.4
tokens/s
138.94
sec
pp384+tg1152
9
tokens/s
3.0
tokens/s
41.07
sec
pp64+tg1024
10
tokens/s
3.1
tokens/s
6.67
sec
pp16+tg1536
10
tokens/s
3.0
tokens/s
1.96
sec