TEST #3409 RESULTS
04/28/2026 - 9:02 PM
ACCELERATOR
MODEL
55.3
tokens/s
generation
2.77
sec
time to first token
2235
tokens/s
prompt
355
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core i5-10400 CPU @ 2.90GHz (skylake)
RAM
31.2GB
OS
Linux
Kernel Release
6.17.0-22-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #22~24.04.1-Ubuntu SMP PREEMPT_DYNAMIC Thu Mar 26 15:25:54 UTC 2
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2762
tokens/s
61.5
tokens/s
387
ms
pp4096+tg256
1959
tokens/s
39.8
tokens/s
2.12
sec
pp2048+tg256
2410
tokens/s
51.4
tokens/s
869
ms
pp2048+tg768
2418
tokens/s
49.2
tokens/s
866
ms
pp1024+tg1024
2745
tokens/s
56.3
tokens/s
389
ms
pp1280+tg3072
2553
tokens/s
46.9
tokens/s
520
ms
pp384+tg1152
3104
tokens/s
61.8
tokens/s
138
ms
pp64+tg1024
2159
tokens/s
66.6
tokens/s
43
ms
pp16+tg1536
1
tokens/s
64.4
tokens/s
19.58
sec