Search AI Products and News

ByteDance's AI code editor, Trae, announces support for OpenAI's latest o3 model, offering developers a more powerful intelligent programming experience. The o3 model, known for its exceptional logical reasoning and tool utilization capabilities, significantly improves the accuracy of code generation and debugging. With the integration of o3, Trae now features intelligent code generation, context-aware debugging, and multi-tool collaboration, enabling it to generate high-quality code based on simple descriptions and accurately identify errors. The developer community has responded enthusiastically, stating that it greatly enhances programming efficiency. Trae also places emphasis on user privacy.

Manus AI launches Wide Research, enabling multi-agent collaboration for complex tasks like data analysis. This major update, available to premium users ($199/month), features cluster technology coordinating dozens of agents. Founder Ji Yichao will demo the 2-month-developed core feature.....

The intelligent agent economy is reshaping the future business ecosystem, with its core being a new economic form built on multi-agent collaborative architecture (IA). Three pillars support this transformation: Decentralized Identifiers (DID) ensure the credible existence of agents as digital citizens; standardized communication protocols (such as MCP) enable seamless collaboration among agents across domains; data containers combined with privacy computing technologies ensure secure data flow. Blockchain and smart contracts form the underlying trust system, driving the creation of a decentralized value network. Agent factories achieve scalable production of agents through modular manufacturing.

Alibaba Cloud launched 'Wuying AgentBay' at WAIC, a cloud platform for AI agent development. It supports code execution, web browsing, and advanced features like vision/NLP, compatible with multiple OS. Developers can access with 3 lines of code, leveraging cloud computing power without local hardware. Offers sandbox, scalable resources, secure storage, and GPU support for efficient AI training.....

Kunlun Wildfire releases the open-source multimodal pre-training model Skywork UniPic, integrating three functions: image understanding, text-to-image generation, and image editing. The model uses an autoregressive training approach, trained on large-scale high-quality data, and has strong generalization and transfer capabilities. Its open-source nature facilitates secondary innovation by developers and can be widely applied in content creation and industry scenarios. This is a significant move by Kunlun Wildfire in the AI field, which will promote the development and popularization of multimodal technology.

PPIO launched China's first Agentic AI infrastructure service platform at the 2025 World Artificial Intelligence Conference, including a general edition and an enterprise edition. The general edition offers a cost-effective distributed GPU cloud foundation and an Agent sandbox compatible with E2B interface, reducing development costs; the enterprise edition is tailored for large enterprises, offering functions such as intelligent decision-making and controllable processes. The platform ensures code reliability through secure sandbox technology, reducing operational costs by 50%, and supports ultra-long context output. PPIO has successfully developed smart education solutions for international schools.

Tencent unveiled the 1+3+NAI comprehensive system at the 2025 World Artificial Intelligence Conference, with the Hunyuan large model as its core, launching innovative technologies such as the Hunyuan 3D World Model and the embodied intelligence platform Tairos. The system includes three platforms: an intelligent agent development platform for B/C ends, an embodied intelligence platform in the robotics field, and an AI application matrix covering multiple industries, comprehensively showcasing Tencent's AI strategic upgrade. By driving with core models and combining with diverse applications, it provides a complete solution for AI industrialization.

NetEase Lingdong launched the world's first embodied intelligent model for excavators in open-pit mines, Lingju, at the World Artificial Intelligence Conference. This model adopts an end-to-end integrated design, breaking traditional development models and achieving three technological breakthroughs: innovative technical pathways, data-driven by real-world scenarios, and full domestication. In field tests in Inner Mongolia coal mines, its loading efficiency has reached 80% of human levels and it is adaptable to harsh environments. NetEase also announced the release of an open-source dataset and the initiation of the 2027 Industry Collaboration Plan, aiming to work with industry partners to promote the unmanned operation of over 30 mines. The core technology, the mechanical intelligence framework Zhixin, supports multiple functions.

Google experts stated at Search Central Live that AI won't disrupt SEO fundamentals. AI functions within traditional search frameworks, keeping core rules unchanged. While SEO techniques remain the same, AI summaries reduce click-through rates for regular results, challenging website visibility and diminishing traditional sites' influence in AI-driven search ecosystems.....

GitHub's AI tool Spark enables rapid web app creation via natural language, reducing development time from hours to minutes. It supports multi-iteration, deployment, and cross-platform use, currently available for Copilot Pro+ users.....

Hedra introduces Hedra Live Avatars: low-cost ($0.05/min), ultra-low latency (<100ms), AI-compatible video generation from static images. Ideal for content, education, customer service, and gaming. Free trial available, with multi-modal upgrades planned.....

Amazon acquires AI wearable firm Bee, known for its $49.99 smart band that generates reminders via audio. This move highlights Amazon's expansion into affordable AI wearables, though privacy concerns persist due to its data management history. Bee's team will join Amazon post-acquisition.....

Dia Browser introduces Agent Mode with 'AI Cursor' for parallel human-AI tasks like auto-searching. Local-first privacy, natural language commands. Redefines browsing interaction.....

A Chinese team has released the open-source AI operating system NeuralOS, which realizes the GUI concept for the AI era proposed by Karpathy. The system uses two modules, RNN and neural renderer, to predict and simulate the Windows operating interface in real-time, accurately displaying user operation feedback. The development team trained the system using a large number of operation videos, and it can currently accurately predict user operations, but still has limitations in handling fast keyboard input. NeuralOS already offers an online trial version, showcasing a new experience of an AI-generated operating system. As the code...

Zhiyuan Robotics secures strategic investment from CP Group, marking a new phase in Chinese robotics' global expansion. The partnership aims to enhance tech innovation and access international markets, focusing on life sciences and retail sectors.....

Mistral releases open-source audio model Voxtral, offering cost-effective AI solutions with multilingual support. Three versions available, starting at $0.001/min. Outperforms Whisper at lower cost.....

UTCP, as an alternative to MCP, directly connects tool endpoints via JSON-defined functions, eliminating proxy layers for lower latency while maintaining security. Its simplicity and compatibility spark developer interest as a potential AI tool standard.....

Meitu launches RoboNeo, an AI tool for image editing, brand design, and web creation via natural language commands. It automates tasks like wedding photo retouching, brand kits, and e-commerce content, reducing visual production barriers for SMEs. Initial tests show rapid multi-format output generation within 5 minutes, though lighting details need refinement.....