Artificial Analysis Openness Index

Name: Artificial Analysis Openness Index
Published: 2025-12-01

Indexes · 2025-12-01

The Artificial Analysis Openness Index summarizes model openness and accessibility factors such as weight availability, license, and transparency. Higher index scores indicate greater openness.

Top models (higher is better)

Model	Score
Apertus 70B Instruct	88.9
Apertus 8B Instruct	88.9
K2 Think V2	88.9
K2-V2-Instruct	88.9
Molmo 7B-D	88.9
OLMo 2 32B Instruct	88.9
OLMo 2 1124 7B Instruct	88.9
Olmo 3.1 32B Instruct	88.9
Olmo 3.1 32B Think	88.9
Olmo 3 32B Think	88.9
Olmo 3 7B Instruct	88.9
Olmo 3 7B Think	88.9
MiniCPM5-1B	83.3
Nemotron 3.5 Lightning 30B A3B (precision unspecified)	83.3
Nemotron 3.5 Super VL 120B A12B (precision unspecified)	83.3
Nemotron 3 Nano 30B A3B	83.3
Nemotron 3 Nano 4B	83.3
Nemotron 3 Super 120B A12B	83.3
Nemotron 3 Ultra 550B A55B	83.3
Nemotron Cascade 2 30B A3B	83.3
Granite 4.2 30B	72.2
Granite 4.2 3B	72.2
Granite 4.2 8B	72.2
INTELLECT-3	72.2
Molmo 2 8B	72.2
Nemotron 3.5 Bolt 74B-A9B (precision unspecified)	72.2
Nemotron Nano 12B V2 VL	72.2
Nemotron Nano 9B V2	72.2
MiniCPM5-2B	66.7
Granite 4.1 30B	61.1
Granite 4.1 3B	61.1
Granite 4.1 8B	61.1
Kimi Linear 48B A3B Instruct	61.1
ERNIE 4.5 300B A47B	55.6
GLM 4.5 Air	55.6
GLM-4.5	55.6
Granite 4.0 1B	55.6
Granite 4.0 350M	55.6
Granite 4.0 H 1B	55.6
Granite 4.0 H 350M	55.6