TEST #597 RESULTS
04/08/2025 - 12:25 PM
ACCELERATOR
MODEL
31.6
tokens/s
generation
4.09
sec
time to first token
312
tokens/s
prompt
134
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M1 Max 8P+2E
RAM
32GB
OS
Darwin
Kernel Release
24.5.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 24.5.0: Wed Mar 26 22:22:16 PDT 2025; root:xnu-11417.120.80.501.1~2/RELEASE_ARM64_T6000
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
446
tokens/s
39.5
tokens/s
2.32
sec
pp4096+tg256
306
tokens/s
25.6
tokens/s
13.43
sec
pp2048+tg256
340
tokens/s
30.5
tokens/s
6.06
sec
pp2048+tg768
324
tokens/s
28.9
tokens/s
6.35
sec
pp1024+tg1024
331
tokens/s
31.1
tokens/s
3.12
sec
pp1280+tg3072
324
tokens/s
28.0
tokens/s
3.99
sec
pp384+tg1152
344
tokens/s
33.1
tokens/s
1.14
sec
pp64+tg1024
296
tokens/s
34.5
tokens/s
239
ms
pp16+tg1536
100
tokens/s
33.3
tokens/s
184
ms