TEST #2270 RESULTS
01/13/2026 - 2:20 AM
ACCELERATOR
MODEL
71.9
tokens/s
generation
633
ms
time to first token
1952
tokens/s
prompt
605
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Xeon W-2133 CPU @ 3.60GHz (skylake-avx512)
RAM
94.3GB
OS
Linux
Kernel Release
6.8.0-90-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #91-Ubuntu SMP PREEMPT_DYNAMIC Tue Nov 18 14:14:30 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1873
tokens/s
77.8
tokens/s
560
ms
pp4096+tg256
2011
tokens/s
62.1
tokens/s
2.05
sec
pp2048+tg256
2311
tokens/s
71.6
tokens/s
900
ms
pp2048+tg768
2317
tokens/s
69.9
tokens/s
898
ms
pp1024+tg1024
2442
tokens/s
74.8
tokens/s
432
ms
pp1280+tg3072
2334
tokens/s
63.9
tokens/s
562
ms
pp384+tg1152
2114
tokens/s
74.8
tokens/s
194
ms
pp64+tg1024
1711
tokens/s
77.9
tokens/s
49
ms
pp16+tg1536
455
tokens/s
73.8
tokens/s
47
ms