TEST #2321 RESULTS
01/17/2026 - 9:03 AM
ACCELERATOR
MODEL
5.5
tokens/s
generation
36.24
sec
time to first token
37
tokens/s
prompt
18
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 5800XT 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
40
tokens/s
5.2
tokens/s
26.08
sec
pp4096+tg256
35
tokens/s
4.9
tokens/s
118.01
sec
pp2048+tg256
37
tokens/s
5.6
tokens/s
56.11
sec
pp2048+tg768
40
tokens/s
5.6
tokens/s
51.98
sec
pp1024+tg1024
38
tokens/s
5.8
tokens/s
27.10
sec
pp1280+tg3072
38
tokens/s
5.4
tokens/s
34.27
sec
pp384+tg1152
39
tokens/s
5.6
tokens/s
10.10
sec
pp64+tg1024
38
tokens/s
5.9
tokens/s
1.87
sec
pp16+tg1536
33
tokens/s
5.7
tokens/s
651
ms