TEST #3301 RESULTS
04/13/2026 - 6:34 PM
ACCELERATOR
MODEL
348
tokens/s
generation
92
ms
time to first token
14452
tokens/s
prompt
3794
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
31.2GB
OS
Linux
Kernel Release
6.18.7-76061807-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #202601231045~1769703228~24.04~cb87b5b SMP PREEMPT_DYNAMIC Thu J
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
18260
tokens/s
366
tokens/s
59
ms
pp4096+tg256
12473
tokens/s
297
tokens/s
332
ms
pp2048+tg256
16065
tokens/s
339
tokens/s
131
ms
pp2048+tg768
16114
tokens/s
333
tokens/s
130
ms
pp1024+tg1024
18370
tokens/s
352
tokens/s
58
ms
pp1280+tg3072
17473
tokens/s
326
tokens/s
76
ms
pp384+tg1152
19992
tokens/s
367
tokens/s
22
ms
pp64+tg1024
9731
tokens/s
382
tokens/s
9
ms
pp16+tg1536
1589
tokens/s
374
tokens/s
12
ms