TEST #2672 RESULTS
02/15/2026 - 6:07 PM
ACCELERATOR
MODEL
63.8
tokens/s
generation
624
ms
time to first token
2084
tokens/s
prompt
597
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 5 5600X 6-Core Processor (znver3)
RAM
31.9GB
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
2465
tokens/s
66.4
tokens/s
431
ms
pp4096+tg256
1962
tokens/s
59.6
tokens/s
2.10
sec
pp2048+tg256
2270
tokens/s
64.5
tokens/s
918
ms
pp2048+tg768
2279
tokens/s
62.5
tokens/s
914
ms
pp1024+tg1024
2458
tokens/s
65.5
tokens/s
432
ms
pp1280+tg3072
2367
tokens/s
58.7
tokens/s
556
ms
pp384+tg1152
2540
tokens/s
65.8
tokens/s
167
ms
pp64+tg1024
1820
tokens/s
65.9
tokens/s
50
ms
pp16+tg1536
591
tokens/s
65.7
tokens/s
42
ms