TEST #2029 RESULTS
12/22/2025 - 8:18 PM
ACCELERATOR
MODEL
53.9
tokens/s
generation
635
ms
time to first token
2052
tokens/s
prompt
558
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
12th Gen Intel Core i7-12700H (alderlake)
RAM
31.8GB
OS
Windows
Kernel Release
10.0
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64
RUNTIME
Name
llamafile
Version
0.9.3
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2459
tokens/s
54.7
tokens/s
436
ms
pp4096+tg256
1950
tokens/s
52.3
tokens/s
2.12
sec
pp2048+tg256
2249
tokens/s
54.3
tokens/s
932
ms
pp2048+tg768
2241
tokens/s
55.9
tokens/s
937
ms
pp1024+tg1024
2416
tokens/s
54.2
tokens/s
446
ms
pp1280+tg3072
2341
tokens/s
52.5
tokens/s
569
ms
pp384+tg1152
2544
tokens/s
52.5
tokens/s
172
ms
pp64+tg1024
1823
tokens/s
53.1
tokens/s
52
ms
pp16+tg1536
448
tokens/s
55.5
tokens/s
56
ms