TEST #2250 RESULTS
01/12/2026 - 11:43 AM
ACCELERATOR
MODEL
66.1
tokens/s
generation
599
ms
time to first token
2204
tokens/s
prompt
624
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
QEMU Virtual CPU version 2.5+
RAM
7.8GB
OS
Linux
Kernel Release
6.8.0-90-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #91-Ubuntu SMP PREEMPT_DYNAMIC Tue Nov 18 14:14:30 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2580
tokens/s
67.9
tokens/s
412
ms
pp4096+tg256
2022
tokens/s
60.6
tokens/s
2.04
sec
pp2048+tg256
2371
tokens/s
65.3
tokens/s
879
ms
pp2048+tg768
2367
tokens/s
64.7
tokens/s
881
ms
pp1024+tg1024
2581
tokens/s
67.0
tokens/s
412
ms
pp1280+tg3072
2498
tokens/s
63.5
tokens/s
527
ms
pp384+tg1152
2678
tokens/s
68.2
tokens/s
158
ms
pp64+tg1024
2112
tokens/s
69.2
tokens/s
44
ms
pp16+tg1536
628
tokens/s
68.7
tokens/s
39
ms