TEST #2294 RESULTS
01/15/2026 - 5:47 AM
ACCELERATOR
MODEL
34.5
tokens/s
generation
1.33
sec
time to first token
979
tokens/s
prompt
294
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core i5-8500T CPU @ 2.10GHz (skylake)
RAM
15.6GB
OS
Linux
Kernel Release
6.8.0-90-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #91-Ubuntu SMP PREEMPT_DYNAMIC Tue Nov 18 14:14:30 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1001
tokens/s
35.5
tokens/s
1.05
sec
pp4096+tg256
908
tokens/s
30.1
tokens/s
4.54
sec
pp2048+tg256
1084
tokens/s
34.6
tokens/s
1.92
sec
pp2048+tg768
1095
tokens/s
34.6
tokens/s
1.90
sec
pp1024+tg1024
1203
tokens/s
35.6
tokens/s
878
ms
pp1280+tg3072
1147
tokens/s
32.7
tokens/s
1.15
sec
pp384+tg1152
1217
tokens/s
35.4
tokens/s
342
ms
pp64+tg1024
863
tokens/s
36.0
tokens/s
100
ms
pp16+tg1536
297
tokens/s
35.7
tokens/s
79
ms