TEST #1727 RESULTS
11/03/2025 - 10:48 PM
ACCELERATOR
MODEL
51.7
tokens/s
generation
1.03
sec
time to first token
1216
tokens/s
prompt
394
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 3700X 8-Core Processor (znver2)
RAM
31.3GB
OS
Linux
Kernel Release
6.12.48+deb13-amd64
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #1 SMP PREEMPT_DYNAMIC Debian 6.12.48-1 (2025-09-20)
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1392
tokens/s
54.5
tokens/s
753
ms
pp4096+tg256
1226
tokens/s
44.0
tokens/s
3.36
sec
pp2048+tg256
1362
tokens/s
50.5
tokens/s
1.52
sec
pp2048+tg768
1360
tokens/s
49.7
tokens/s
1.53
sec
pp1024+tg1024
1426
tokens/s
52.8
tokens/s
735
ms
pp1280+tg3072
1393
tokens/s
47.7
tokens/s
937
ms
pp384+tg1152
1396
tokens/s
54.7
tokens/s
292
ms
pp64+tg1024
1032
tokens/s
56.3
tokens/s
78
ms
pp16+tg1536
361
tokens/s
55.3
tokens/s
60
ms