TEST #3154 RESULTS
03/22/2026 - 8:43 PM
ACCELERATOR
MODEL
212
tokens/s
generation
105
ms
time to first token
12769
tokens/s
prompt
2953
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 9 5900X 12-Core Processor (znver3)
RAM
127.9GB
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
15686
tokens/s
225
tokens/s
70
ms
pp4096+tg256
11244
tokens/s
211
tokens/s
370
ms
pp2048+tg256
14934
tokens/s
219
tokens/s
142
ms
pp2048+tg768
14001
tokens/s
207
tokens/s
152
ms
pp1024+tg1024
16689
tokens/s
211
tokens/s
66
ms
pp1280+tg3072
15676
tokens/s
207
tokens/s
89
ms
pp384+tg1152
17177
tokens/s
202
tokens/s
29
ms
pp64+tg1024
7933
tokens/s
210
tokens/s
13
ms
pp16+tg1536
1581
tokens/s
212
tokens/s
14
ms