TEST #2945 RESULTS
03/02/2026 - 5:39 AM
ACCELERATOR
MODEL
34.9
tokens/s
generation
681
ms
time to first token
2027
tokens/s
prompt
470
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
RAM
121.7GB
OS
Linux
Kernel Release
6.17.0-1008-nvidia
Architecture
aarch64
Version
Cosmopolitan 3.9.7 MODE=aarch64; #8-Ubuntu SMP PREEMPT_DYNAMIC Wed Jan 21 17:56:56 UTC 2026
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2464
tokens/s
36.4
tokens/s
443
ms
pp4096+tg256
1708
tokens/s
32.0
tokens/s
2.44
sec
pp2048+tg256
2160
tokens/s
34.7
tokens/s
980
ms
pp2048+tg768
2159
tokens/s
34.1
tokens/s
976
ms
pp1024+tg1024
2500
tokens/s
35.1
tokens/s
438
ms
pp1280+tg3072
2371
tokens/s
33.1
tokens/s
569
ms
pp384+tg1152
2916
tokens/s
35.9
tokens/s
159
ms
pp64+tg1024
1499
tokens/s
36.5
tokens/s
70
ms
pp16+tg1536
469
tokens/s
36.2
tokens/s
61
ms