TEST #3369 RESULTS
04/23/2026 - 1:32 PM
ACCELERATOR
MODEL
179
tokens/s
generation
318
ms
time to first token
3654
tokens/s
prompt
1272
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M4 Max 12P+4E
RAM
64GB
OS
Darwin
Kernel Release
25.4.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 25.4.0: Thu Mar 19 19:33:25 PDT 2026; root:xnu-12377.101.15~1/RELEASE_ARM64_T6041
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
5028
tokens/s
206
tokens/s
209
ms
pp4096+tg256
4169
tokens/s
123
tokens/s
991
ms
pp2048+tg256
4397
tokens/s
161
tokens/s
472
ms
pp2048+tg768
4373
tokens/s
148
tokens/s
476
ms
pp1024+tg1024
4561
tokens/s
171
tokens/s
230
ms
pp1280+tg3072
4397
tokens/s
149
tokens/s
298
ms
pp384+tg1152
3911
tokens/s
209
tokens/s
102
ms
pp64+tg1024
1653
tokens/s
228
tokens/s
43
ms
pp16+tg1536
397
tokens/s
218
tokens/s
44
ms