TEST #2510 RESULTS
02/01/2026 - 11:38 PM
ACCELERATOR
MODEL
29.9
tokens/s
generation
1.55
sec
time to first token
1151
tokens/s
prompt
281
LocalScore
HOW YOU STACK UP
Explore All ResultsKomdigiUB 8B Instruct DTP - Q4_K - Medium
SYSTEM
CPU
Intel Core i5-14450HX (alderlake)
RAM
15.7GB
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
1361
tokens/s
30.8
tokens/s
785
ms
pp4096+tg256
536
tokens/s
27.3
tokens/s
7.68
sec
pp2048+tg256
1256
tokens/s
29.8
tokens/s
1.66
sec
pp2048+tg768
1248
tokens/s
29.6
tokens/s
1.68
sec
pp1024+tg1024
1432
tokens/s
30.2
tokens/s
748
ms
pp1280+tg3072
1319
tokens/s
28.9
tokens/s
1.00
sec
pp384+tg1152
1577
tokens/s
30.8
tokens/s
275
ms
pp64+tg1024
1266
tokens/s
31.1
tokens/s
82
ms
pp16+tg1536
369
tokens/s
30.9
tokens/s
75
ms