TEST #2539 RESULTS
02/05/2026 - 4:35 AM
ACCELERATOR
MODEL
21.9
tokens/s
generation
6.18
sec
time to first token
202
tokens/s
prompt
89
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M1 Pro 8P+2E
RAM
32GB
OS
Darwin
Kernel Release
25.2.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 25.2.0: Tue Nov 18 21:09:40 PST 2025; root:xnu-12377.61.12~1/RELEASE_ARM64_T6000
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
223
tokens/s
24.0
tokens/s
4.64
sec
pp4096+tg256
211
tokens/s
16.8
tokens/s
19.45
sec
pp2048+tg256
220
tokens/s
20.6
tokens/s
9.37
sec
pp2048+tg768
220
tokens/s
19.9
tokens/s
9.34
sec
pp1024+tg1024
216
tokens/s
22.3
tokens/s
4.78
sec
pp1280+tg3072
228
tokens/s
19.6
tokens/s
5.66
sec
pp384+tg1152
226
tokens/s
24.1
tokens/s
1.74
sec
pp64+tg1024
195
tokens/s
25.1
tokens/s
366
ms
pp16+tg1536
75
tokens/s
24.6
tokens/s
251
ms