Robots Atlas>ROBOTS ATLAS
JoyAI-VL-Interaction
AI Models

JoyAI-VL-Interaction

A JD (JoyAI) vision-language multimodal model focused on interaction — combining vision and language for interactive and agentic tasks.
✓ Active⏳ Limited accessMultimodalVision
🔬JD Explore AcademyResearch lab
Access:APIDeployment:☁ Cloud

Overview

JoyAI-VL-Interaction is a vision-language multimodal model developed within JD's research (the JoyAI brand, including JD Explore Academy).

The model combines image and natural-language understanding with an emphasis on interaction — scene understanding, image-grounded dialogue and tasks that link visual perception to natural-language instructions, useful in e-commerce, customer service and agentic applications.

Detailed technical parameters are provided by JD.

Classification
MultimodalVision
Access & deployment
API
Cloud
Weights: Closed
Key parameters
📥 Input: image, text

Technical specification

Modalities
⬇ Input
imagetext
⬆ Output
text

Sources and related pages