Welcome to the "AI Daily" section! This is your guide to exploring the world of artificial intelligence every day. Every day, we present you with the latest content in the AI field, focusing on developers, helping you understand technology trends and learn about innovative AI product applications.
Fresh AI products click to learn more:https://top.aibase.com/
1. ByteDance announced the official open source of two core projects of Coze: Coze Studio and Coze Loop
ByteDance has open-sourced the core projects of its AI Agent development platform Coze, Coze Studio and Coze Loop, providing developers with low-barrier and highly flexible tools. This move marks its strategic layout in the AI field and promotes industry innovation.
【AiBase Summary:】
🚀 Coze Studio is a visual AI Agent development platform that supports drag-and-drop workflow building, suitable for fast development and deployment of AI applications.
🔧 Coze Loop focuses on the full lifecycle management of AI Agents, including prompt debugging and performance evaluation, helping optimize agent performance.
🌐 Both projects are under the Apache 2.0 license, allowing free use, modification, and even commercial deployment, lowering the entry barrier.
More details: https://github.com/coze-dev/coze-studio
2. Keling AI officially launched the Ling Animation Canvas Creative Workbench, with significant upgrades to the Multi-Image Reference feature
Keling AI announced two important developments at the 2025 World Artificial Intelligence Conference: launching the new creative workbench 'Ling Animation Canvas' and upgrading the 'Multi-Image Reference' feature. At the same time, the number of global creators on Keling AI has exceeded 45 million, showing the widespread application of AI-generated content and the increasing market recognition.
【AiBase Summary:】
🎨 Ling Animation Canvas provides infinite visual space, intelligent creation assistance, and real-time collaboration functions, improving creative efficiency.
🖼️ After the upgrade, the multi-image reference function has significantly improved in character consistency, scene uniformity, and style retention.
📈 The user count of Keling AI has exceeded 45 million, covering 149 countries and regions globally, with rapid growth in B-end customers.
3. Runway launches a new video editing model Aleph
Runway company launched a new video editing model Aleph, known as the "Kontext model for the video field." This model enables diverse video editing through natural language instructions, including content addition and deletion, style transfer, green screen keying, which greatly lowers the video creation barrier and provides content creators with unprecedented freedom.
【AiBase Summary:】
🎥 Supports diverse video editing through natural language instructions, such as adding or removing content and style transfer.
🔄 Enables environment and atmosphere transformation, such as changing a sunny scene to a rainy one or adjusting it to night view.
💡 Built-in green screen keying function and support for re-lighting, improving post-production efficiency.
More details: https://runwayml.com/research/introducing-runway-aleph
4.金山办公发布WPS AI 3.0灵犀版本 打造原生Office智能体
Kingsoft Office officially released WPS AI 3.0 version - WPS Lixi at the 2025 AI Productivity Forum, marking the upgrade of office AI from a tool to an intelligent assistant. This version uses a native Office intelligent body design, supporting natural language multi-turn dialogue, realizing document creation, presentation generation, and voice assistant functions. At the same time, the addition of the WPS knowledge base and AI search function further enriches the office intelligent ecosystem.
【AiBase Summary:】
🧠 WPS Lixi supports multi-turn dialogue, realizing seamless interaction between document creation and presentation generation.
📊 Lixi retains the original format, ensuring that AI-generated content meets users' actual needs.
🌐 The integration of WPS knowledge base and AI search function provides comprehensive intelligent office solutions.
5. JD.com announces its large model brand has been upgraded to JoyAI and releases the new intelligent brand JoyInside
JD.com announced at the 2025 World Artificial Intelligence Conference that it has upgraded its large model brand to JoyAI and launched a new intelligent brand JoyInside, demonstrating its strong strength and innovation capability in the field of artificial intelligence.
【AiBase Summary:】
🧠 JoyAI large model series covers multiple modalities to meet different user needs.
💡 Introduce innovative technologies to improve reasoning efficiency and reduce training costs.
🤖 JoyInside is widely applied, covering humanoid robots, children's toys, and other carriers.
6. Pudong Launches 2 Billion Yuan Artificial Intelligence Seed Fund, First Phase 500 Million Focused on Early Stage Investment
Pudong launched a 2 billion yuan artificial intelligence seed fund, with the first phase of 500 million yuan focused on early-stage investment, aiming to solve the financing difficulties of start-up enterprises and promote the development of the artificial intelligence industry.
【AiBase Summary:】
🧠 The Pudong Artificial Intelligence Seed Fund has a total scale of 2 billion yuan, with the first phase of 500 million yuan focused on seed round investment.
🚀 The scale of the Pudong AI industry exceeds 160 billion yuan, accounting for 40% of the city, forming a significant cluster effect.
📈 Five measures to promote the new AI ecology, including scenario expansion, infrastructure, and financial support.
7. Alibaba Open Sources WebSailor AI Agent, Breaking Information Retrieval Barriers
The article introduces the open-source AI agent framework WebSailor launched by Alibaba's Tongyi Lab, which excels in handling complex tasks and improves information retrieval efficiency through reinforcement learning and complex task generation technology. Its open-source nature provides developers with more exploration space.
【AiBase Summary:】
🌐 WebSailor is an open-source AI agent framework, focusing on complex task processing.
🧠 Combining reinforcement learning with complex task generation enhances model performance.
🔍 Information blurring technology enhances task challenge, improving model intelligence level.
More details: https://github.com/Alibaba-NLP/WebAgent
8. China's Large Model Count Exceeds 1500, Global Leadership Highlighted
The article showcases China's significant achievements in the field of artificial intelligence, including the leading number of large models globally, the widespread application of generative AI products, and the rapid growth of the industry scale. It also mentions the deep application of domestic AI products in multiple industries and the accelerated internationalization process, indicating that China's competitiveness in the global AI field is continuously increasing.
【AiBase Summary:】
🌍 There are 3,755 large models globally, with 1,509 contributed by Chinese companies, leading the world.
📈 Generative AI product user ratio reaches 80.9%, and the industry scale continues to grow.
🚀 High-quality domestic AI products are accelerating their overseas expansion, significantly enhancing China's international influence.
9. Ant Digital Collaborates with Multiple Organizations to Release Finova Large Model Financial Application Evaluation Benchmark
Ant Digital officially released the financial inference large model Agentar-Fin-R1 at the World Artificial Intelligence Conference forum. The model shows excellent performance in financial professionalism, reasoning ability, and security compliance. By building a comprehensive financial task data system and innovative training algorithms, it enhances the model's ability to handle complex tasks and collaborates with multiple organizations to launch the Finova evaluation benchmark, promoting the industry to improve the application level of large models in the financial field.
【AiBase Summary:】
🧠 Agentar-Fin-R1 performs exceptionally well in the financial field, surpassing similar models.
📊 A comprehensive financial task classification system was built, covering multiple financial scenarios.
🚀 The Finova evaluation benchmark is open-sourced, helping the industry improve the application level of large models.
10. Tencent Hunyuan 3D World Model Shockingly Released! 360° Immersive Scene Free Experience, AI Reimagines the Future of the Virtual World!
Tencent released the Hunyuan 3D World Model 1.0 at the 2025 World Artificial Intelligence Conference and fully open-sourced it. The model supports high-precision 360° scene generation and interactive roaming features, greatly lowering the barrier to 3D content creation, bringing new possibilities to game development, VR, and digital content creation fields.
【AiBase Summary:】
📷 High-precision 360° scene generation, enhancing visual experience.
🕹️ Interactive roaming scenes, smart range judgment optimizing experience.
🌐 Fully open-sourced, empowering the global developer community.