JoyAI

JoyAI is an Artificial Intelligence brand introduced by JD.com, unveiled at the 2025 World artificial intelligence Conference. Rather than a single product, it bundles a family of models and APPlications covering text, voice, image, vIDEo and Digital Human generation. The web edition focuses on accessible AIGC creation, making it a prACTical entry point for users who want to produce visuals and short videos without a production team.

What it does

The web platform puts three creative engines in one place: Text-to-Image, Text-to-Video and digital human generation. Users can describe a scene and receive high-DeFinition imagery, turn an idea into a short video CLIp, or build a customizable virtual presenter with a chosen appearance and voice for talking-head content. Around these core tools sit extras such as an AI toolbox and podcast-style features, and the brand connects to JD.com's wider hardware and assistant ecosystem.

Key capabilities

• Image Generation: Prompt-driven creation of commercial-grade still visuals.
• Video Generation: quick clips and popular short-video formats from simple inputs.
• digital humans: adjustable avatars and voices for narration and broadcast-style videos.
• Model range: a reported family of models from small to very large sizes, tuned for different speed and quality trade-offs.
• Beyond creation: an open Agent platform, an embodied-intelligence layer for devices, and a real-time video-language interaction model aimed at live monitoring and commentary.

Who it is for

JoyAI suits content creators, small merchants and marketers who need fast promotional visuals or presenter videos, as well as teams exploring AI Agents and hardware integration. developers interested in open Frameworks and datasets may also find the agent and embodied-AI components relevant, though those are more technical than the consumer web tools.

How to get started

Visit the official website in a desktop browser, sign in with the available account options, and pick a creation mode such as image, video or digital human. Describe what you want in the prompt field, adjust any style or voice settings offered, then generate and review the result. The interface is primarily in Chinese, so non-Chinese speakers may need translation assistance.

Pricing and access

The platform is presented as freely accessible, and the core generation tools can be tried without an obvious upfront payment. Credit allowances, queue priority, watermark rules and paid tiers may change and may vary by region, so confirm current terms on the site before relying on it for commercial work. Access appears open without a VPN, although availability outside mainland China is not guaranteed.