TEST #2502 RESULTS
02/01/2026 - 11:12 PM
ACCELERATOR
MODEL
60.0
tokens/s
generation
568
ms
time to first token
2518
tokens/s
prompt
643
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 3B Instruct - 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
2853
tokens/s
62.2
tokens/s
374
ms
pp4096+tg256
2022
tokens/s
51.8
tokens/s
2.04
sec
pp2048+tg256
2518
tokens/s
59.3
tokens/s
830
ms
pp2048+tg768
2542
tokens/s
58.7
tokens/s
822
ms
pp1024+tg1024
2930
tokens/s
61.3
tokens/s
366
ms
pp1280+tg3072
2800
tokens/s
56.6
tokens/s
474
ms
pp384+tg1152
3483
tokens/s
62.8
tokens/s
126
ms
pp64+tg1024
2664
tokens/s
63.7
tokens/s
40
ms
pp16+tg1536
847
tokens/s
63.7
tokens/s
34
ms