TEST #4081 RESULTS
08/23/2026 - 5:37 PM
ACCELERATOR
MODEL
56.1
tokens/s
generation
2.83
sec
time to first token
2271
tokens/s
prompt
356
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core i5-10400F CPU @ 2.90GHz (skylake)
RAM
23.3GB
OS
Linux
Kernel Release
7.0.0-30-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #30~24.04.1-Ubuntu SMP PREEMPT_DYNAMIC Fri Aug 7 13:27:52 UTC 2
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2560
tokens/s
59.3
tokens/s
418
ms
pp4096+tg256
1974
tokens/s
40.7
tokens/s
2.10
sec
pp2048+tg256
2444
tokens/s
52.8
tokens/s
857
ms
pp2048+tg768
2444
tokens/s
51.0
tokens/s
857
ms
pp1024+tg1024
2802
tokens/s
58.0
tokens/s
381
ms
pp1280+tg3072
2691
tokens/s
48.0
tokens/s
492
ms
pp384+tg1152
3229
tokens/s
63.0
tokens/s
134
ms
pp64+tg1024
2296
tokens/s
66.7
tokens/s
42
ms
pp16+tg1536
1
tokens/s
65.5
tokens/s
20.22
sec