TEST #3758 RESULTS
06/19/2026 - 12:20 AM
ACCELERATOR
MODEL
69.6
tokens/s
generation
544
ms
time to first token
2435
tokens/s
prompt
678
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core i9-10900X CPU @ 3.70GHz (cascadelake)
RAM
30.3GB
OS
Linux
Kernel Release
7.0.0-22-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #22-Ubuntu SMP PREEMPT_DYNAMIC Mon May 25 15:54:34 UTC 2026
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2867
tokens/s
71.7
tokens/s
371
ms
pp4096+tg256
2216
tokens/s
61.5
tokens/s
1.87
sec
pp2048+tg256
2604
tokens/s
67.2
tokens/s
801
ms
pp2048+tg768
2601
tokens/s
66.9
tokens/s
802
ms
pp1024+tg1024
2868
tokens/s
70.3
tokens/s
371
ms
pp1280+tg3072
2793
tokens/s
66.6
tokens/s
472
ms
pp384+tg1152
3120
tokens/s
73.5
tokens/s
136
ms
pp64+tg1024
2092
tokens/s
74.9
tokens/s
44
ms
pp16+tg1536
757
tokens/s
73.8
tokens/s
34
ms