Skip to main content
Modelmultimodalv1.0

Cambrian-1

by New York University (NYU) · free · Last verified 2026-03-17

Cambrian-1 is a research vision-language model from NYU focused on spatial intelligence and visual grounding, introducing the Spatial Vision Aggregator to fuse features from multiple vision encoders. It achieves strong performance on spatial reasoning and visual understanding benchmarks, providing a fully open research platform for multimodal model development.

https://cambrian-mllm.github.io
C
CBelow Average
Adoption: DQuality: B+Freshness: ACitations: CEngagement: F

Specifications

License
Apache 2.0
Pricing
free
Capabilities
vision, visual-question-answering, spatial-reasoning, image-understanding, visual-grounding
Integrations
Hugging Face
Use Cases
spatial-intelligence-research, visual-qa, multimodal-research, robotics-perception
API Available
No
Parameters
34B
Context Window
8K
Modalities
text, image
Training Cutoff
2024
Tags
nyu, research, vision-language, spatial-intelligence, open-source
Added
2026-03-17
Completeness
95%

Index Score

41.8
Adoption
35
Quality
79
Freshness
80
Citations
48
Engagement
0

Explore the full AI ecosystem on Agents as a Service