TEST #3048 RESULTS
03/09/2026 - 1:16 AM
ACCELERATOR
MODEL
218
tokens/s
generation
163
ms
time to first token
8204
tokens/s
prompt
2223
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Xeon CPU E5-2650 v2 @ 2.60GHz (ivybridge)
RAM
122.8GB
OS
Linux
Kernel Release
6.14.0-37-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #37-Ubuntu SMP PREEMPT_DYNAMIC Fri Nov 14 22:10:32 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
10200
tokens/s
238
tokens/s
105
ms
pp4096+tg256
7026
tokens/s
169
tokens/s
589
ms
pp2048+tg256
9115
tokens/s
207
tokens/s
230
ms
pp2048+tg768
9111
tokens/s
201
tokens/s
230
ms
pp1024+tg1024
10480
tokens/s
223
tokens/s
102
ms
pp1280+tg3072
9903
tokens/s
193
tokens/s
134
ms
pp384+tg1152
11207
tokens/s
241
tokens/s
38
ms
pp64+tg1024
5953
tokens/s
249
tokens/s
15
ms
pp16+tg1536
844
tokens/s
243
tokens/s
23
ms