TEST #2258 RESULTS
01/12/2026 - 4:28 PM
ACCELERATOR
MODEL
33.7
tokens/s
generation
649
ms
time to first token
1951
tokens/s
prompt
466
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
13th Gen Intel Core i5-1335U (alderlake)
RAM
63.2GB
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
2332
tokens/s
34.7
tokens/s
468
ms
pp4096+tg256
1963
tokens/s
31.0
tokens/s
2.12
sec
pp2048+tg256
2242
tokens/s
33.6
tokens/s
943
ms
pp2048+tg768
2231
tokens/s
32.7
tokens/s
947
ms
pp1024+tg1024
2425
tokens/s
34.1
tokens/s
451
ms
pp1280+tg3072
2325
tokens/s
32.1
tokens/s
579
ms
pp384+tg1152
2466
tokens/s
34.8
tokens/s
185
ms
pp64+tg1024
1173
tokens/s
35.0
tokens/s
82
ms
pp16+tg1536
402
tokens/s
34.7
tokens/s
66
ms