Back to Directory
LLaVA
AI Model
Low Confidence
LLaVA represents a novel end-to-end trained large multimodal model that combines a vision encoder and Vicuna for general-purpose visual and language understanding, achieving impressive chat capabilities mimicking spirits of the multimodal GPT-4 and setting a new state-of-the-art accuracy on Science QA
Transparency Score
Weighted across four pillars · updated May 22, 2026
Overall grade
0% weighted
Model
—
Not yet assessed
Vendor
—
Not yet assessed
Infrastructure
—
Not yet assessed
Capabilities
Image-Text-To-Text
Help us complete this listing
We don't have vendor details for LLaVA yet. Know the provider's website?
Compliance & Risk
Get insights into risk by running assessments on this AI application.
Added: May 22, 2026
Updated: May 22, 2026
Ready to manage AI applications?
Track, assess, and govern your AI applications with Anove.