TEST #2303 RESULTS
01/15/2026 - 2:31 PM
ACCELERATOR
MODEL
65.9
tokens/s
generation
605
ms
time to first token
2150
tokens/s
prompt
616
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
11th Gen Intel Core i7-11700K @ 3.60GHz (rocketlake)
RAM
7.8GB
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
2594
tokens/s
68.3
tokens/s
410
ms
pp4096+tg256
2026
tokens/s
60.0
tokens/s
2.04
sec
pp2048+tg256
2333
tokens/s
64.8
tokens/s
893
ms
pp2048+tg768
2335
tokens/s
64.5
tokens/s
892
ms
pp1024+tg1024
2511
tokens/s
66.7
tokens/s
423
ms
pp1280+tg3072
2494
tokens/s
63.4
tokens/s
528
ms
pp384+tg1152
2514
tokens/s
68.1
tokens/s
167
ms
pp64+tg1024
2095
tokens/s
69.0
tokens/s
45
ms
pp16+tg1536
450
tokens/s
68.5
tokens/s
50
ms