TEST #1078 RESULTS
06/19/2025 - 7:17 PM
ACCELERATOR
MODEL
72.9
tokens/s
generation
768
ms
time to first token
2052
tokens/s
prompt
580
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Xeon W-2150B CPU @ 3.00GHz (skylake-avx512)
RAM
251.3GB
OS
Linux
Kernel Release
6.8.12-11-pve
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #1 SMP PREEMPT_DYNAMIC PMX 6.8.12-11 (2025-05-22T09:39Z)
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2430
tokens/s
93.2
tokens/s
432
ms
pp4096+tg256
1355
tokens/s
11.0
tokens/s
3.11
sec
pp2048+tg256
1921
tokens/s
67.1
tokens/s
1.08
sec
pp2048+tg768
1913
tokens/s
62.9
tokens/s
1.09
sec
pp1024+tg1024
2406
tokens/s
78.0
tokens/s
437
ms
pp1280+tg3072
2274
tokens/s
41.4
tokens/s
575
ms
pp384+tg1152
3144
tokens/s
93.3
tokens/s
131
ms
pp64+tg1024
2020
tokens/s
108
tokens/s
40
ms
pp16+tg1536
1008
tokens/s
101
tokens/s
24
ms