TEST #2538 RESULTS
02/04/2026 - 11:07 PM
ACCELERATOR
MODEL
30.6
tokens/s
generation
2.72
sec
time to first token
487
tokens/s
prompt
176
LocalScore
HOW YOU STACK UP
Explore All ResultsGranite 3.1 8b Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 5700G with Radeon Graphics (znver3)
RAM
27.3GB
OS
Linux
Kernel Release
6.17.12-300.fc43.x86_64
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #1 SMP PREEMPT_DYNAMIC Sat Dec 13 05:06:24 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.3
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
582
tokens/s
37.0
tokens/s
1.79
sec
pp4096+tg256
435
tokens/s
6.8
tokens/s
9.56
sec
pp2048+tg256
523
tokens/s
27.9
tokens/s
3.95
sec
pp2048+tg768
522
tokens/s
26.5
tokens/s
3.96
sec
pp1024+tg1024
580
tokens/s
32.4
tokens/s
1.79
sec
pp1280+tg3072
551
tokens/s
20.1
tokens/s
2.35
sec
pp384+tg1152
590
tokens/s
38.5
tokens/s
673
ms
pp64+tg1024
224
tokens/s
44.3
tokens/s
305
ms
pp16+tg1536
376
tokens/s
41.6
tokens/s
62
ms