TEST #3085 RESULTS
03/15/2026 - 1:18 PM
ACCELERATOR
MODEL
48.2
tokens/s
generation
391
ms
time to first token
3231
tokens/s
prompt
736
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 7800X3D 8-Core Processor (znver4)
RAM
61.9GB
OS
Linux
Kernel Release
6.18.7-76061807-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #202601231045~1769703228~24.04~cb87b5b SMP PREEMPT_DYNAMIC Thu J
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
4219
tokens/s
53.5
tokens/s
262
ms
pp4096+tg256
3243
tokens/s
39.2
tokens/s
1.29
sec
pp2048+tg256
3811
tokens/s
46.4
tokens/s
559
ms
pp2048+tg768
3827
tokens/s
45.1
tokens/s
556
ms
pp1024+tg1024
4222
tokens/s
48.9
tokens/s
262
ms
pp1280+tg3072
4022
tokens/s
43.6
tokens/s
338
ms
pp384+tg1152
3862
tokens/s
51.7
tokens/s
120
ms
pp64+tg1024
1387
tokens/s
53.3
tokens/s
69
ms
pp16+tg1536
489
tokens/s
52.3
tokens/s
63
ms