TEST #3091 RESULTS
03/15/2026 - 2:20 PM
ACCELERATOR
MODEL
52.2
tokens/s
generation
907
ms
time to first token
1429
tokens/s
prompt
435
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
11th Gen Intel Core i7-11700 @ 2.50GHz (rocketlake)
RAM
2.9GB
OS
Linux
Kernel Release
6.12.73+deb13-amd64
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #1 SMP PREEMPT_DYNAMIC Debian 6.12.73-1 (2026-02-17)
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1457
tokens/s
54.2
tokens/s
721
ms
pp4096+tg256
1356
tokens/s
45.9
tokens/s
3.04
sec
pp2048+tg256
1577
tokens/s
51.2
tokens/s
1.32
sec
pp2048+tg768
1580
tokens/s
50.5
tokens/s
1.32
sec
pp1024+tg1024
1716
tokens/s
53.0
tokens/s
615
ms
pp1280+tg3072
1672
tokens/s
49.3
tokens/s
784
ms
pp384+tg1152
1795
tokens/s
54.7
tokens/s
232
ms
pp64+tg1024
1420
tokens/s
55.9
tokens/s
62
ms
pp16+tg1536
286
tokens/s
55.3
tokens/s
73
ms