TEST #2994 RESULTS
03/06/2026 - 8:57 PM
ACCELERATOR
MODEL
59.6
tokens/s
generation
4.82
sec
time to first token
298
tokens/s
prompt
154
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core i9-14900K (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
307
tokens/s
62.1
tokens/s
3.35
sec
pp4096+tg256
249
tokens/s
51.7
tokens/s
16.48
sec
pp2048+tg256
284
tokens/s
57.8
tokens/s
7.22
sec
pp2048+tg768
280
tokens/s
57.9
tokens/s
7.33
sec
pp1024+tg1024
308
tokens/s
60.6
tokens/s
3.33
sec
pp1280+tg3072
306
tokens/s
57.1
tokens/s
4.20
sec
pp384+tg1152
333
tokens/s
62.5
tokens/s
1.17
sec
pp64+tg1024
334
tokens/s
63.8
tokens/s
206
ms
pp16+tg1536
275
tokens/s
63.0
tokens/s
73
ms