TEST #1467 RESULTS
09/01/2025 - 3:26 PM
ACCELERATOR
MODEL
61.8
tokens/s
generation
1.22
sec
time to first token
1248
tokens/s
prompt
399
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 9 9950X3D 16-Core Processor
RAM
59.4GB
OS
Linux
Kernel Release
6.14.0-29-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #29-Ubuntu SMP PREEMPT_DYNAMIC Thu Aug 7 18:32:38 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1435
tokens/s
64.9
tokens/s
730
ms
pp4096+tg256
866
tokens/s
54.4
tokens/s
4.75
sec
pp2048+tg256
1199
tokens/s
60.5
tokens/s
1.73
sec
pp2048+tg768
1201
tokens/s
59.5
tokens/s
1.72
sec
pp1024+tg1024
1457
tokens/s
62.3
tokens/s
721
ms
pp1280+tg3072
1408
tokens/s
58.3
tokens/s
926
ms
pp384+tg1152
1566
tokens/s
65.1
tokens/s
261
ms
pp64+tg1024
1320
tokens/s
65.8
tokens/s
64
ms
pp16+tg1536
775
tokens/s
65.0
tokens/s
37
ms