TEST #1653 RESULTS
10/14/2025 - 6:34 PM
ACCELERATOR
MODEL
361
tokens/s
generation
60
ms
time to first token
24736
tokens/s
prompt
5287
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
QEMU Virtual CPU version 2.5+
RAM
251.4GB
OS
Linux
Kernel Release
5.14.0-570.42.2.el9_6.x86_64
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #1 SMP PREEMPT_DYNAMIC Sun Sep 14 13:59:34 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
36374
tokens/s
375
tokens/s
31
ms
pp4096+tg256
16749
tokens/s
319
tokens/s
248
ms
pp2048+tg256
25820
tokens/s
345
tokens/s
82
ms
pp2048+tg768
25937
tokens/s
345
tokens/s
82
ms
pp1024+tg1024
36146
tokens/s
363
tokens/s
31
ms
pp1280+tg3072
33192
tokens/s
340
tokens/s
41
ms
pp384+tg1152
33364
tokens/s
381
tokens/s
14
ms
pp64+tg1024
11396
tokens/s
398
tokens/s
8
ms
pp16+tg1536
3648
tokens/s
389
tokens/s
7
ms