TEST #3202 RESULTS
03/28/2026 - 2:15 PM
ACCELERATOR
MODEL
48.8
tokens/s
generation
383
ms
time to first token
3305
tokens/s
prompt
750
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 7800X3D 8-Core Processor (znver4)
RAM
61.9GB
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
2924
tokens/s
52.8
tokens/s
368
ms
pp4096+tg256
3308
tokens/s
39.9
tokens/s
1.26
sec
pp2048+tg256
4086
tokens/s
47.1
tokens/s
522
ms
pp2048+tg768
4097
tokens/s
46.1
tokens/s
521
ms
pp1024+tg1024
4625
tokens/s
49.7
tokens/s
240
ms
pp1280+tg3072
4375
tokens/s
44.6
tokens/s
313
ms
pp384+tg1152
4447
tokens/s
52.4
tokens/s
106
ms
pp64+tg1024
1446
tokens/s
53.9
tokens/s
61
ms
pp16+tg1536
438
tokens/s
52.9
tokens/s
55
ms