TEST #3420 RESULTS
04/29/2026 - 7:37 PM
ACCELERATOR
MODEL
29.5
tokens/s
generation
4.22
sec
time to first token
298
tokens/s
prompt
128
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M4 Pro 8P+4E
RAM
24GB
OS
Darwin
Kernel Release
25.3.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 25.3.0: Wed Jan 28 20:51:28 PST 2026; root:xnu-12377.91.3~2/RELEASE_ARM64_T6041
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
334
tokens/s
32.0
tokens/s
3.09
sec
pp4096+tg256
307
tokens/s
23.2
tokens/s
13.38
sec
pp2048+tg256
326
tokens/s
28.2
tokens/s
6.32
sec
pp2048+tg768
325
tokens/s
27.3
tokens/s
6.35
sec
pp1024+tg1024
332
tokens/s
30.2
tokens/s
3.11
sec
pp1280+tg3072
317
tokens/s
26.3
tokens/s
4.07
sec
pp384+tg1152
315
tokens/s
32.0
tokens/s
1.25
sec
pp64+tg1024
298
tokens/s
33.6
tokens/s
243
ms
pp16+tg1536
124
tokens/s
32.8
tokens/s
157
ms