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 to help you understand technical trends and innovative AI product applications.
Fresh AI products Click to learn more:https://app.aibase.com/zh
1. Doubao 2.1Pro launched on Huoshan Fangzhou, Feishu 8.0 introduces team AI agent "Doubao Work Partner"
The productivity products of ByteDance launched a major update, with Huoshan Yingqing upgrading Doubao's large model to the 0915 version and optimizing API access; Feishu released version 8.0, introducing the team AI agent "Doubao Work Partner," promoting the transformation of AI from personal assistant to team collaboration.
AiBase Highlights:
🧠 Model upgrades enhance evidence tracing and data verification, improving stability for complex tasks.
💻 Enhanced multimodal programming capabilities, supporting code repository analysis and repair testing.
👥 Introduction of the team AI agent "Doubao Work Partner," enabling AI collaboration and sharing.
2. Shengshu Technology officially launched Vidu S2, supporting real-time editing, dynamic reference images, and headset compatibility
Shengshu Technology officially launched the Vidu S2 video large model series, including two core models: Avatar and Editing, and explored VR headset space video generation and editing capabilities, demonstrating strong technical strength and application potential.
AiBase Highlights:
🎥 Vidu S2-Avatar supports voice control and dynamic reference images, achieving high interactive freedom in digital character creation.
📹 Vidu S2-Editing enables real-time video stream editing, supporting style transfer, virtual try-on, character replacement, and background replacement.
🕶️ Shengshu Technology explores VR headset space video generation and editing capabilities, enhancing immersive experiences.
3. Claude released a major upgrade: 43 all-around workflows and 27 major applications deeply integrated
Anthropic made significant functional upgrades to its 'Small and Medium Business Version Claude,' launching 43 efficient workflows and deep integration with 27 mainstream business software, aiming to help small and medium enterprises use artificial intelligence as a core business driver.
AiBase Highlights:
🧠 43 efficient workflows help small and medium enterprises improve efficiency.
🤝 Deep integration with 27 mainstream business software, building a complete ecosystem matrix.
🔒 Strengthened security and data privacy governance to ensure enterprise data security.
4. Google launches a multilingual AI new strategy, making every language in the world heard and understood
Google released a new multilingual AI strategy aimed at making every language in the world heard and understood. The strategy includes multiple technological achievements, such as support for over 300 languages, real-time spoken translation, support for low-resource languages, and accessibility design.
AiBase Highlights:
🌐 Google launched multilingual AI technology, supporting 7.1 billion people and 86% of the population.
🎙️ Promoting native audio intelligence, achieving real-time spoken translation across 70 languages.
🌍 Through the "1,000 Languages Program" and open-source data collaboration, it enhances support for low-resource languages.
5. Chinese Internet Basic Corpus 4.0 officially released, 120GB high-quality trusted data重磅上线
The release of Chinese Internet Basic Corpus 4.0 marks an important progress in China's artificial intelligence field, providing solid data support for large model training and the high-quality development of the industry.
AiBase Highlights:
🧠 Release of Chinese Internet Basic Corpus 4.0, providing data support for large model training.
🌐 Guided by the Central Cyberspace Administration, multiple industry units jointly promoted the project.
📊 120GB high-quality trusted data was launched, helping the development of the AI industry.
6. OpenAI internal Lily project exposed, are human reviewers browsing users' private chat records?
The article revealed OpenAI's internal "Lily Project," which involves human reviewers assessing user chat records for quality, while also raising concerns about privacy and security.
AiBase Highlights:
🧠 Human reviewers assess ChatGPT user chat records for quality.
🔒 There is a risk of privacy leaks during the review process, which may include user personal information.
🔄 OpenAI did not clearly inform users that their chat records might be reviewed by humans.
7. vivo Blue Heart Large Model fully upgraded: Four models launched, capable of listening, remembering, and completing tasks
vivo announced the full upgrade of the Blue Heart Large Model at the 2026 Developer Conference, launching four core models covering speech understanding, edge-side perception and memory, and cloud-side task execution, showcasing its leading position in the field of intelligent agent large models.
AiBase Highlights:
🧠 The Blue LM-RealTime voice large model has strong environmental adaptability, capable of identifying complex scenarios such as noisy environments, accents, and mixed Chinese and English speech input.
💾 The Blue LM-Nano edge-side large model can perceive and store personal memories in real time, integrating and processing various types of information, and recording user habits and preferences over a long period.
⚙️ The Blue LM-Flash and Blue LM-Pro cloud-side large models, through self-developed Agentic engines and deep collaboration with system-level Harness, act as the central control for mobile phone tasks, enabling cross-device and cross-application task execution.
8. Google officially announced Gemini 3.8 Live, pioneering the "speaking and thinking at the same time" revolutionary interaction in the industry
Google announced the launch of Gemini 3.8 Live and its extended version, achieving a major breakthrough in the field of voice interaction, especially realizing deep reasoning capabilities in low-latency voice streams, bringing a brand-new transformation to voice AI applications.
AiBase Highlights:
🧠 Gemini 3.8 Live realizes deep multi-step reasoning capabilities in low-latency voice streams, changing the application paradigm of voice AI.
💡 Gemini 3.8 Live Extended Thinking provides "thinking while speaking" background extended reasoning capabilities, supporting complex task processing.
🚀 The new generation of Live series models performs excellently in high-intelligence-demand scenarios such as real-time step debugging and complex mathematical reasoning.