Incident: On the evening of January 30, 2024, the company released its 2023 performance forecast. It is expected to achieve net profit of 75 million yuan to 10,000 yuan in 2023, an increase of 81.80%-142.39% year on year; realized deducted non-net profit of 78 million yuan to 105 million yuan, an increase of 636.06 percent to 890.86% year on year, and the annual results achieved high growth, in line with market expectations.
The Big Canopy Model was officially released, and nearly 100 atomic capabilities have been iterated. On January 30, 2024, the Wanxing Tianmu Big Model was officially released at the “Xiangyue AI Promise - Multimedia Big Model Innovation Application Summit Forum and Launch and Signing Ceremony”. “Tianmu” focuses on digital creative creative production scenarios. Based on 1.5 billion creators and 10 billion localized high-quality audio and video data accumulation, it is based on audio and video generative AI technology. It has the three characteristics of “multimedia,” “vertical solutions,” and “localized data.” The whole chain empowers global creators, making large-scale model applications more targeted and effective. Currently, the “Tianmu” model has iterated on nearly 100 audio and video atomic capabilities, and has highlighted multimedia capabilities including Wensheng themed videos, Wensheng 3D videos, AI singers, video AI soundtracks, and digital human broadcasts.
Wanxing AI's open platform has opened an API, and the industry is working together to build a new ecosystem of model applications. The Wanxing AI Open Platform uses the Tianmu Big Model as the technical foundation to provide one-stop AI solutions in the field of digital creativity, involving multiple AI generation and editing capabilities for images, audio, and video, and provides full-scene, easy-to-use, and discounted visual intelligence API services for visual intelligence technology companies and developers (including developers). On January 30, 2024, the company announced that the Wanxing Tianmu Audio and Video Multimedia Big Model R&D Center was officially located in Malanshan; reached a strategic cooperation with Zhongguang Tianze, a state-owned listed company in Changsha Radio and Television, to further expand computational resources; and reached computing power cooperation with HUAWEI CLOUD and Ma Investment Power (Changsha Malanshan Investment, Development and Construction Co., Ltd.) to provide computing power support for big model training. All sectors joined forces to build a new AI ecosystem.
Investment advice: The company fully embraces AIGC. The self-developed canopy model has been officially released, and the multi-modal generation and editing products are expected to be further updated and iterated. We expect the company to achieve operating income of 1,497/18.97/2,499 billion yuan in 2023-2025, and net profit to mother of 0.886/171/250 million yuan. The corresponding PS is 7.8/6.1/4.8 times, respectively, maintaining the “gain” rating.
Risk warning: New product promotion falls short of expectations; implementation of subscription system falls short of expectations; creative software competition pattern worsens.