TEST #3290 RESULTS
04/11/2026 - 3:39 PM
ACCELERATOR
MODEL
47.7
tokens/s
generation
142
ms
time to first token
9439
tokens/s
prompt
1469
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
12215
tokens/s
51.7
tokens/s
103
ms
pp4096+tg256
10869
tokens/s
43.3
tokens/s
400
ms
pp2048+tg256
12260
tokens/s
50.1
tokens/s
187
ms
pp2048+tg768
12252
tokens/s
45.1
tokens/s
188
ms
pp1024+tg1024
12627
tokens/s
49.8
tokens/s
101
ms
pp1280+tg3072
11588
tokens/s
45.1
tokens/s
131
ms
pp384+tg1152
10836
tokens/s
49.2
tokens/s
59
ms
pp64+tg1024
1838
tokens/s
47.3
tokens/s
56
ms
pp16+tg1536
466
tokens/s
47.9
tokens/s
55
ms