TEST #2629 RESULTS
02/14/2026 - 2:19 PM
ACCELERATOR
MODEL
59.2
tokens/s
generation
678
ms
time to first token
1905
tokens/s
prompt
550
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 5800X 8-Core Processor (znver3)
RAM
63.9GB
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
2273
tokens/s
60.2
tokens/s
467
ms
pp4096+tg256
1808
tokens/s
56.0
tokens/s
2.28
sec
pp2048+tg256
2093
tokens/s
59.6
tokens/s
996
ms
pp2048+tg768
2089
tokens/s
59.3
tokens/s
998
ms
pp1024+tg1024
2266
tokens/s
59.7
tokens/s
469
ms
pp1280+tg3072
2181
tokens/s
58.1
tokens/s
604
ms
pp384+tg1152
2350
tokens/s
60.0
tokens/s
180
ms
pp64+tg1024
1636
tokens/s
60.1
tokens/s
56
ms
pp16+tg1536
449
tokens/s
59.9
tokens/s
52
ms