TEST #3289 RESULTS
04/11/2026 - 3:20 PM
ACCELERATOR
MODEL
46.7
tokens/s
generation
143
ms
time to first token
9444
tokens/s
prompt
1456
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 7800X3D 8-Core Processor (znver4)
RAM
31.1GB
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
12348
tokens/s
50.0
tokens/s
102
ms
pp4096+tg256
10781
tokens/s
42.6
tokens/s
405
ms
pp2048+tg256
11983
tokens/s
48.3
tokens/s
192
ms
pp2048+tg768
12265
tokens/s
44.4
tokens/s
187
ms
pp1024+tg1024
13285
tokens/s
48.9
tokens/s
97
ms
pp1280+tg3072
11216
tokens/s
44.5
tokens/s
135
ms
pp384+tg1152
10825
tokens/s
48.3
tokens/s
60
ms
pp64+tg1024
1828
tokens/s
46.6
tokens/s
55
ms
pp16+tg1536
466
tokens/s
47.2
tokens/s
55
ms