TEST #2533 RESULTS
02/04/2026 - 9:47 AM
ACCELERATOR
MODEL
106
tokens/s
generation
74
ms
time to first token
18081
tokens/s
prompt
2962
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 9800X3D 8-Core Processor
RAM
60.4GB
OS
Linux
Kernel Release
6.18.8-1-default
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #1 SMP PREEMPT_DYNAMIC Fri Jan 30 12:17:19 UTC 2026 (0df318e)
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
25258
tokens/s
103
tokens/s
50
ms
pp4096+tg256
19749
tokens/s
96.6
tokens/s
218
ms
pp2048+tg256
22638
tokens/s
99.8
tokens/s
101
ms
pp2048+tg768
22606
tokens/s
99.2
tokens/s
101
ms
pp1024+tg1024
25081
tokens/s
102
tokens/s
51
ms
pp1280+tg3072
21869
tokens/s
101
tokens/s
69
ms
pp384+tg1152
20059
tokens/s
119
tokens/s
29
ms
pp64+tg1024
4296
tokens/s
114
tokens/s
24
ms
pp16+tg1536
1174
tokens/s
118
tokens/s
20
ms