TEST #3893 RESULTS
07/12/2026 - 12:38 PM
ACCELERATOR
MODEL
39.6
tokens/s
generation
3.22
sec
time to first token
463
tokens/s
prompt
179
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen AI 9 HX 370 w/ Radeon 890M
RAM
30GB
OS
Linux
Kernel Release
7.0.0-22-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #22-Ubuntu SMP PREEMPT_DYNAMIC Mon May 25 15:54:34 UTC 2026
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
486
tokens/s
41.8
tokens/s
2.13
sec
pp4096+tg256
360
tokens/s
34.6
tokens/s
11.42
sec
pp2048+tg256
427
tokens/s
38.6
tokens/s
4.83
sec
pp2048+tg768
429
tokens/s
38.0
tokens/s
4.80
sec
pp1024+tg1024
488
tokens/s
40.0
tokens/s
2.12
sec
pp1280+tg3072
478
tokens/s
37.1
tokens/s
2.71
sec
pp384+tg1152
533
tokens/s
41.4
tokens/s
744
ms
pp64+tg1024
545
tokens/s
42.6
tokens/s
140
ms
pp16+tg1536
421
tokens/s
42.0
tokens/s
60
ms