XIAMEN, China, June 17, 2026 — At the 2026 Meitu Multimedia Festival, Meitu unveiled four new products — Picchi, Artflo, MVLAND, and MeituHub — alongside major upgrades to Zcool, DesignKit, Kaipai, and RoboNeo. Together, the eight offerings collectively drive AI applications from delivering features to delivering outcomes.

8 AI Products Unveiled Across Portrait Editing, Talking Videos, Short Dramas, Design, Music Videos, and More

At the festival, Meitu unveiled eight AI products spanning a wide range of creative scenarios — from designer communities and portrait retouching to commercial design, talking videos, concept visuals, music visualization, AI short dramas, and AI-powered visual production pipelines.

Among the four new offerings, Picchi is a dedicated portrait-retouching Agent that enables users to replicate their aesthetic preferences through features such as "Style DNA / Creator DNA". Artflo focuses on inspiration management and conceptual visual creation. MVLAND leverages multi-Agent collaboration to rapidly generate high-quality visuals for music. MeituHub serves as a technical solution and the foundation of the AI visual production pipeline, connecting models, tools, and workflows.

Meitu also updated four existing products: Zcool further connects designers, brands, and creative demand through industry competitions and workflow enhancements; DesignKit introduces an AI Design Team, shifting from content generation to delivering commercial outcomes; Kaipai launches an AI Assistant that supports the entire workflow from topic planning to finished videos; and RoboNeo introduces an AI Short Drama Team, making short-drama creation more accessible.

With these updates, Meitu's full suite of photo/video products enters a new chapter. In photo, Meitu, BeautyCam, Kumoo, and Picchi form a product matrix. In video, the portfolio includes Wink, Kaipai, RoboNeo, MVLAND, and Artflo. In design, users can access DesignKit and WHEE. MeituHub serves as an open technology platform; Zcool acts as the community and asset library; and the Miraclevision large model underpins the entire ecosystem.

From Agent to Agent Teams, From Delivering Features to Delivering Outcomes

According to Wu Xinhong, Meitu's founder and CEO, the company is evolving across four dimensions: products are shifting from apps to AI Agents; business models, from subscriptions to AI compute credits, organizational structures, from established setups to AI-native innovation teams, and from features to outcomes.

"At last year's Multimedia Festival, we talked about AI workflows, which were still largely feature-centric," said Wu. "This year, we want to go a step further and deliver outcomes directly to users."

Meitu's research found that learning AI applications is not part of most people's core job responsibilities, the iteration speed of AI tools far outpaces what individuals can realistically keep up with, and some users would rather pay for results than become experts themselves.

On the technology front, Meitu's Agent Teams solution enables multiple specialized Agents to divide responsibilities, collaborate, and cross-check outputs across multiple stages. This allows them to execute complex workflows and deliver better business outcomes.

In terms of service models, Meitu has introduced Human-Led Creator Services across multiple products for users who prefer not to learn AI tools. By combining native AI workflows with hands-on expertise from professional designers, these services offer experiences such as "Design for Me" and "Create Videos for Me."

Miraclevision V6 Debuts; Meitu Launches RMB 100 Million Product Challenge

The festival also saw the launch of Miraclevision V6. Built on a Mixture-of-Experts (MoE) architecture, it features advanced visual decision-making capabilities and supports multimodal inputs across text, images, video, and audio. Combined with Meitu's expertise in user understanding, aesthetic preferences, and real-world knowledge, it can accurately interpret user intent and systematically break down visual tasks.

According to Meitu, from January to May 2026, an average of 96.3% of calls to generative AI features across its photo/video products came from Miraclevision. Leveraging its powerful capabilities, Miraclevision V6 will be deeply integrated into Meitu's photo/video products, continuously feeding back into product innovation and accelerating the rollout of new features.

In addition, Meitu has announced a RMB 100 million incubation fund and launched the Meitu Hatch Catch Product Challenge, inviting AI developers worldwide to submit innovative photo/video products. Eligible entries must be AI-native photo/video applications designed for global markets, already launched, and supported by an initial user base.

Selected teams may receive up to RMB 5 million in investment support. They may also have the opportunity to join Meitu as AI innovation studios, gaining deeper resource synergy and collaboration opportunities. Meitu will provide incubation support covering product refinement, user validation, growth resources, and technical exchange, to help products grow and scale.