China AI Native Industry Insights – 20250529 – Tencent | Alibaba | ByteDance | more
Explore Tencent’s release of HunyuanVideo-Avatar for seamless digital human creation, Alibaba and SAP’s strategic collaboration for cloud and AI advancements, and Coze Space’s innovative one-click AI-driven podcast creation. Discover more in Today’s China AI Native Industry Insights.
1. Tencent Open-Sources HunyuanVideo-Avatar for One-Click Digital Human Creation
🔑 Key Details:
– New Technology: HunyuanVideo-Avatar generates videos of talking/singing subjects from a single image and audio clip, maintaining high consistency and natural movements.
– Capability Range: Supports head-and-shoulders, half-body, and full-body videos with lip synchronization and emotional expressions across multiple styles and characters.
– Deployment Status: Single-subject capability is now open-sourced on GitHub and available on Tencent Hunyuan’s website, with more features coming soon.
– Real-world Applications: Already implemented in QQ Music, Kugou Music, and WeSing for AI singers, audiobooks, and personalized music videos.
💡 How It Helps:
– Content Creators: Significantly reduces video production costs and time with one-click generation of talking-head and full-body videos.
– Marketers: Enables quick creation of product introductions and interactive advertisements with natural-looking digital humans.
– App Developers: Open-source model provides foundation for integrating digital human capabilities into music, storytelling, and entertainment applications.
🌟 Why It Matters:
This technology represents a significant advancement in accessibility of digital human creation, bringing professional-grade video generation to the masses. By open-sourcing capabilities that exceed many existing solutions, Tencent is democratizing digital human technology while positioning itself as a leader in multimodal AI that bridges audio, visual, and emotional dimensions.
Original Chinese article: https://mp.weixin.qq.com/s/6YWrkVITRUmjNDyZhXrXLA
English translation via free online service: https://translate.google.com/translate?hl=en&sl=zh-CN&tl=en&u=https%3A%2F%2Fmp.weixin.qq.com%2Fs%2F6YWrkVITRUmjNDyZhXrXLA
Video Credit: Hunyuan (@TencentHunyuan on X)
2. Alibaba and SAP Form Strategic Partnership for Cloud and AI Innovation
🔑 Key Details:
– Strategic Partnership: Alibaba Group and SAP announced collaboration combining SAP’s enterprise software with Alibaba Cloud infrastructure and AI capabilities.
– AI Integration: SAP will explore integrating Alibaba’s Tongyi Qianwen large language model into SAP AI Core’s generative AI Hub.
– Geographic Focus: Partnership initially targets Chinese market before expanding to Southeast Asia, Middle East, and Africa.
– Cloud Deployment: Alibaba Cloud will support deployment of SAP ERP Cloud and private cloud versions for enterprise customers.
💡 How It Helps:
– Enterprise IT Leaders: Access to scalable, secure and intelligent solutions combining SAP software with Alibaba Cloud’s infrastructure.
– Business Strategists: New opportunities to accelerate innovation and digital transformation with integrated AI capabilities.
– Regional Businesses: Better localized AI functionality for SAP applications in target markets through native language model integration.
🌟 Why It Matters:
This collaboration represents a significant advancement in enterprise cloud and AI integration in Asia and emerging markets. By combining SAP’s industry-leading enterprise solutions with Alibaba’s cloud infrastructure and AI models, both companies strengthen their competitive positions against global rivals while addressing specific regional needs. The partnership also signals Alibaba’s continued expansion into enterprise services beyond its e-commerce roots.
Original Chinese article: https://mp.weixin.qq.com/s/32FSP-SXBF1_U3j2W2QbOQ
English translation via free online service: https://translate.google.com/translate?hl=en&sl=zh-CN&tl=en&u=https%3A%2F%2Fmp.weixin.qq.com%2Fs%2F32FSP-SXBF1_U3j2W2QbOQ
Video Credit: The original article
3. Coze Space Launches One-Click Podcast Creation with AI
🔑 Key Details:
– New Feature: Coze Space (coze.cn) now offers one-click AI-powered podcast generation.
– Dual Modes: Users can input audience info for AI-generated original content or convert existing articles into podcast format.
– Enhanced Audio: Features natural voices with emotional tone, dual narrators, and professional podcast structure.
– Integrated Output: Seamlessly combines with Coze tools to generate multimedia pages with embedded audio.
💡 How It Helps:
– Content Creators: Removes the need for complex planning, recording, or editing in podcast production.
– Marketers: Instantly transforms written materials (articles, PDFs, slides) into engaging spoken content.
– News Responders: Enables fast podcast turnaround for timely commentary or trending topics.
– Non-technical Users: Delivers studio-quality results without needing equipment or production skills.
🌟 Why It Matters:
Coze’s AI podcasting tool lowers the barrier to high-quality audio content creation, enabling faster, more scalable production. It empowers individuals and teams to repurpose content or start from scratch—no recording studio required. This shift reflects a broader trend in generative AI, making traditionally complex formats like podcasts accessible to anyone, and bringing more voices into the audio space.
Original Chinese article: https://mp.weixin.qq.com/s/p2h8RT82mILGPl-QrdWYFw
English translation via free online service: https://translate.google.com/translate?hl=en&sl=zh-CN&tl=en&u=https%3A%2F%2Fmp.weixin.qq.com%2Fs%2Fp2h8RT82mILGPl-QrdWYFw
Video Credit: The original article
4. Rednote Partners with Fudan University on AI Humanities Training
🔑 Key Details:
– AI Humanities Camp: Rednote’s hi lab and Fudan University’s Philosophy School launched a joint training program on May 27.
– Post-Training Focus: Aims to enhance human-aligned expression and values as part of AI post-training.
– Team Upgrade: Rednote’s internal AI team rebranded as hi lab (Humane Intelligence Lab), now recruiting humanities trainers.
– Interdisciplinary Collaboration: Encourages students from diverse fields to form teams for a competition starting this September.
💡 How It Helps:
– Philosophy Researchers: Gain a direct role in shaping AI systems with ethical and humanistic insights.
– AI Engineers: Work alongside humanities experts to improve models’ reasoning, empathy, and alignment.
– University Students: Get hands-on experience in the “AI + Humanities” field through structured training and competition.
– Educators: Explore new models for cultivating talent that bridges technology and the humanities.
🌟 Why It Matters:
This initiative marks a notable shift in China’s AI development strategy, emphasizing ethical alignment and human values. By integrating Fudan’s humanities depth with Rednote’s technical resources, the program fosters a new generation of interdisciplinary talent. It reflects growing global awareness that future AI must not only be powerful, but also responsible and empathetic.
Original Chinese article: https://mp.weixin.qq.com/s/bz_AQ2h9fvnfeY-zIKV9Xw
English translation via free online service: https://translate.google.com/translate?hl=en&sl=zh-CN&tl=en&u=https%3A%2F%2Fmp.weixin.qq.com%2Fs%2Fbz_AQ2h9fvnfeY-zIKV9Xw
Video Credit: The original article
5. DeepSeek-R1 Upgrade Rivals Claude 4 in Programming Performance
🔑 Key Details:
– New Model Release: DeepSeek-R1-0528 was open-sourced on Hugging Face, showcasing significant improvements in programming and reasoning abilities.
– Performance Benchmark: Tests show the model approaches OpenAI’s o4 mini and higher versions of o3, making it a top performer among open-source models.
– Text Handling: Shows improved accuracy for responding to questions about materials within 32k context window, though performance drops at 60k.
– Code Generation: Capable of producing 1000+ lines of bug-free code in one generation with enhanced HTML/CSS/JS capabilities.
💡 How It Helps:
– AI Developers: Access to a high-performing open-source model that rivals proprietary options like Claude 4.
– Programmers: Better assistance with complex coding tasks and interactive web development projects.
– Content Creators: Improved accuracy when working with lengthy documents and technical materials.
– Application Builders: Enhanced capabilities for game development and interactive application creation.
🌟 Why It Matters:
This release represents a significant narrowing of the gap between open-source and proprietary AI models. What was potentially planned as DeepSeek’s R2 release appears to have been expedited due to competitive pressure, highlighting the accelerating pace of innovation in the AI space. The model’s ability to match or exceed Claude 4 in certain tasks demonstrates how open-source alternatives are becoming increasingly viable options for advanced AI applications.
Original Chinese article: https://mp.weixin.qq.com/s/lzrhUmzRUN41rIFq606urg
English translation via free online service: https://translate.google.com/translate?hl=en&sl=zh-CN&tl=en&u=https%3A%2F%2Fmp.weixin.qq.com%2Fs%2FlzrhUmzRUN41rIFq606urg
Video Credit: The original article
That’s all for today’s China AI Native Industry Insights. Join us at AI Native Foundation Membership Dashboard for the latest insights on AI Native, or follow our linkedin account at AI Native Foundation and our twitter account at AINativeF.