Search AI Products and News

The French AI lab Mistral has officially launched its first series of reasoning models - Magistral, marking another significant progress in the AI field. This series includes two versions: Magistral Small and Magistral Medium, aiming to enhance logical reasoning capabilities in fields such as mathematics and physics. These models solve problems step-by-step, comparable to models like OpenAI's o3 and Google's Gemini2.5Pro.

The French AI lab Mistral recently announced the official release of its first inference model series -- Magistral. The series includes two versions: Magistral Small and Magistral Medium, aimed at enhancing logical reasoning capabilities in disciplines such as mathematics and physics. Mistral stated that the Magistral series models solve problems step by step to improve the consistency and reliability of results.

Recently, Pixeltorte's independently developed "Fangtang large model" has officially passed the filing by the Cyberspace Administration of China (CAC), becoming the first application-level image large model with official qualifications in the domestic imaging industry. This achievement not only marks a breakthrough in technology for the Fangtang large model but also demonstrates its compliance and safety standards meeting national requirements. The successful registration of Fangtang is an important milestone in the field of artificial intelligence, especially against the backdrop of the increasingly developing image generation technology. Pixeltorte's independent research and development in this technology highlights its strength in AI.

In the continuous advancement of medical technology, United Imaging officially launched its latest Yuanzhi medical large model on April 9th. The launch of this large model not only brings new opportunities to the medical industry but also provides doctors and patients with more intelligent services. United Imaging's Yuanzhi large model integrates tens of millions of medical imaging data and hundreds of thousands of precisely annotated data, aiming to improve the efficiency and accuracy of image diagnosis. The Yuanzhi medical large model supports more than 10 imaging modalities and can handle 300 types of image processing tasks. This means it can...

Recientemente, una nueva herramienta de IA llamada "MagicColor" ha generado un gran debate en los círculos tecnológicos y artísticos. Esta tecnología innovadora permite colorear automáticamente imágenes de bocetos con una simple operación. Solo necesita cargar un boceto en blanco y negro y algunas imágenes de referencia, y el sistema podrá completar un relleno de color preciso y natural en cuestión de segundos. Esta función innovadora no solo aumenta en gran medida la eficiencia de la creación, sino que también proporciona a los diseñadores de anime, ilustradores y usuarios comunes un asistente de creación inteligente sin precedentes. Según la introducción, la función principal de MagicColor radica en su función de un solo clic

Magic Atom's latest release showcases advanced humanoid and quadruped robots, highlighting significant progress in artificial intelligence and robotics, and offering a glimpse into the future of intelligent technology.

Baidu AI has announced the open-sourcing of PP-TableMagic, a new generation table recognition solution, marking a significant breakthrough in structured information extraction from tables. PP-TableMagic aims to address the limitations of traditional table recognition technologies in complex scenarios. Through an innovative multi-model network architecture, it achieves high-precision end-to-end table recognition and supports highly customizable model fine-tuning for all scenarios.

Recently, a tool called Ready AI has been gaining significant attention in the design community. The remarkable feature of this tool is that users can generate professional-level web interfaces in just 30 seconds by inputting text commands. With Ready AI, users have more control over their desired UI. Unlike similar products, Ready AI supports real-time preview and version comparison features, demonstrating incredible efficiency in design testing for social media platforms and e-commerce websites. Ready AI offers various framework options, such as HTML.

Honor officially announced today that its latest DeepSeek-R1 online version has been officially launched. The first batch of supported models for this online version includes the Honor Magic7 series as well as the Honor Magic V3, Vs3, V2, and Vs2 series, showcasing Honor's continuous leadership in technological innovation. As the first manufacturer on the Android platform to integrate DeepSeek, Honor has once again taken the lead by launching the DeepSeek-R1 online version. This new version introduces 'online search.'

Recently, the Honor YOYO Smart Agent Store announced the official launch of the DeepSeek-R1 trial version, bringing a brand new intelligent experience to Honor phone users. As long as users have their Honor phone system version at MagicOS 8.0 or above and upgrade the YOYO Assistant to version 80.0.1.503 or above, they can be among the first to experience this feature.