TEST #2692 RESULTS
02/16/2026 - 10:48 PM
ACCELERATOR
MODEL
81.1
tokens/s
generation
1000
ms
time to first token
1198
tokens/s
prompt
460
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M1 Pro 8P+2E
RAM
16GB
OS
Darwin
Kernel Release
24.6.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 24.6.0: Mon Jul 14 11:30:29 PDT 2025; root:xnu-11417.140.69~1/RELEASE_ARM64_T6000
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1464
tokens/s
91.6
tokens/s
710
ms
pp4096+tg256
1259
tokens/s
53.8
tokens/s
3.27
sec
pp2048+tg256
1413
tokens/s
73.3
tokens/s
1.46
sec
pp2048+tg768
1415
tokens/s
70.2
tokens/s
1.46
sec
pp1024+tg1024
1453
tokens/s
82.7
tokens/s
716
ms
pp1280+tg3072
1437
tokens/s
65.8
tokens/s
902
ms
pp384+tg1152
1271
tokens/s
92.8
tokens/s
312
ms
pp64+tg1024
839
tokens/s
102
tokens/s
85
ms
pp16+tg1536
233
tokens/s
97.2
tokens/s
77
ms