Alibaba Group, a Chinese technology company, has introduced Tongyi Wanxiang, an AI-powered image generator designed to compete with platforms such as Midjourney, Stable Diffusion’s DreamStudio, and OpenAI’s DALL-E.
Tongyi Wanxiang operates by receiving user prompts in either Chinese or English and producing images in various styles. Currently in beta form and available for enterprise users in China, the tool arrives at a time when US companies like Amazon, Google, and Shutterstock are investing heavily in AI-based image generation technology.
Jingren Zhou, CTO of Alibaba Cloud Intelligence, expressed confidence that Tongyi Wanxiang would drive the company’s pursuit of advanced generative AI models as it explores technologies that foster greater creativity and productivity.
In a press release, Zhou stated, “With the introduction of Tongyi Wanxiang, businesses across multiple sectors, including e-commerce, gaming, design, and advertising, will have easier access to high-quality generative AI imagery. This will facilitate the development of innovative AI art and creative expressions.”
What Capabilities Does Tongyi Wanxiang Possess?
When prompted, Tongyi Wanxiang can generate intricate images in various styles, ranging from watercolor and oil paintings to Chinese brushwork animations, sketches, flat illustrations, and 3D cartoons.
Furthermore, this generative AI model can transform images into new ones with similar styles and apply the visual style of one picture to another while maintaining the original content.
Alibaba Cloud’s knowledge arrangement, visual AI, and natural language processing (NLP) technologies power the image generator. It leverages multilingual inputs for enhanced training, boosting its semantic understanding and resulting in more accurate and contextually relevant image generation.
Coinciding with its launch, Alibaba also unveiled ModelScopeGPT, an adaptable framework designed to assist users in accomplishing complex and specialized AI tasks involving language, vision, and speech using various AI models.
In April, the company from Shenzhen announced Tongyi Qianwen, its large language model, in response to the growing popularity of OpenAI’s ChatGPT.
A report by consulting firm McKinsey estimated in June that generative AI has the potential to contribute $4.4 trillion in value to the global economy annually.