TEST #2756 RESULTS
02/19/2026 - 6:58 AM
ACCELERATOR
MODEL
36.5
tokens/s
generation
656
ms
time to first token
2102
tokens/s
prompt
489
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
RAM
119.6GB
OS
Linux
Kernel Release
6.17.0-1008-nvidia
Architecture
aarch64
Version
Cosmopolitan 3.9.7 MODE=aarch64; #8-Ubuntu SMP PREEMPT_DYNAMIC Wed Jan 21 17:56:56 UTC 2026
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2598
tokens/s
36.6
tokens/s
422
ms
pp4096+tg256
1775
tokens/s
33.5
tokens/s
2.34
sec
pp2048+tg256
2219
tokens/s
36.1
tokens/s
952
ms
pp2048+tg768
2253
tokens/s
35.9
tokens/s
938
ms
pp1024+tg1024
2597
tokens/s
37.1
tokens/s
422
ms
pp1280+tg3072
2454
tokens/s
35.3
tokens/s
549
ms
pp384+tg1152
2962
tokens/s
37.8
tokens/s
155
ms
pp64+tg1024
1581
tokens/s
38.3
tokens/s
67
ms
pp16+tg1536
481
tokens/s
38.0
tokens/s
60
ms