TEST #1017 RESULTS
06/07/2025 - 5:16 PM
ACCELERATOR
MODEL
42.5
tokens/s
generation
5.31
sec
time to first token
273
tokens/s
prompt
130
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 PRO 5750G with Radeon Graphics (znver3)
RAM
32GB
OS
Windows
Kernel Release
10.0
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
339
tokens/s
46.1
tokens/s
3.05
sec
pp4096+tg256
211
tokens/s
35.5
tokens/s
19.48
sec
pp2048+tg256
272
tokens/s
41.5
tokens/s
7.57
sec
pp2048+tg768
262
tokens/s
40.4
tokens/s
7.84
sec
pp1024+tg1024
329
tokens/s
43.2
tokens/s
3.14
sec
pp1280+tg3072
258
tokens/s
38.8
tokens/s
4.98
sec
pp384+tg1152
277
tokens/s
45.1
tokens/s
1.41
sec
pp64+tg1024
270
tokens/s
46.7
tokens/s
257
ms
pp16+tg1536
239
tokens/s
45.5
tokens/s
87
ms