According to foreign media reports, informed sources revealed that Microsoft's engineering team is accelerating the deployment of server resources to prepare for OpenAI's upcoming GPT-4.5 and GPT-5 models. Although OpenAI CEO Sam Altman recently confirmed that GPT-4.5 will be unveiled in a few weeks, Microsoft internally expects to complete the hosting deployment of the new model as soon as next week. This move indicates that the technological race in the field of generative AI has entered the sprinting stage. Code-named Orion, GPT-4

In the AI industry, Together AI recently announced the completion of a $305 million Series B funding round, which has garnered widespread attention. The company's rise is closely related to its newly launched deep reasoning model, DeepSeek-R1. Contrary to initial concerns, many industry experts believe that advancements in deep reasoning have not diminished the demand for infrastructure; rather, they are continuously enhancing this demand. Image source note: Image generated by AI, image licensed from service provider Midjourney since its establishment in 2023.

DingTalk AI Corporate Search, launched in 2024, focuses on addressing the pain points of enterprise users related to information fragmentation and low retrieval efficiency. This tool leverages the understanding, reasoning, and generating capabilities of large models to integrate unstructured data such as chat logs, documents, knowledge bases, schedules, and logs into a structured knowledge network, improving search efficiency by 300% over traditional methods. By offering this free access, DingTalk aims to further assist enterprise users in building dynamic knowledge networks and optimizing workflows.

Recently, the successful launch of the VLM-R1 project has brought new hope to this field. This project represents the successful migration of the R1 Method from the DeepSeek team into visual language models, indicating that AI's understanding of visual content will enter a whole new phase. The inspiration for VLM-R1 comes from last year's open-source R1 Method by DeepSeek, which leverages GRPO (Generative Reward Processing Optimization) reinforcement.

In today's fast-paced information age, the efficiency of team communication is particularly important. HiveNexus has launched HiveChat, an AI chat application designed specifically for small to medium-sized teams, aimed at helping teams improve collaboration efficiency and easily meet daily communication needs. The application supports various AI models such as Deepseek, OpenAI, Claude, and Gemini, providing users with a more intelligent and personalized chat experience.

Recently, SuperMap Software announced on its interactive platform that it has successfully completed the integration and adaptation of the DeepSeek-V3 and R1 series large models. This development marks another breakthrough for SuperMap Software in the application of intelligent technology. As a result of this innovation, the SuperMap Copilot Beta version has officially been launched and made available to customers. This product not only fully demonstrates the powerful capabilities of large models but also possesses significant advantages in terms of 'diversity, speed, quality, and cost-effectiveness,' showcasing SuperMap Software's commitment to advancing smart production.

Meta has recently announced that it will hold the first LlamaCon developer conference on April 29, focusing on the latest advancements in its Llama series of generative AI models. The event aims to provide developers with a platform to share the latest achievements in open-source AI, helping them to better build applications and products. As generative AI technology rapidly evolves, Meta hopes to further promote the development of its ecosystem through this conference. At the upcoming event, Meta plans to showcase several.

On Monday local time, Mistral unveiled a large model called Mistral Saba in Paris, which is specifically optimized for Arabic interaction capabilities. This innovative move is seen as a significant breakthrough in the European AI field. The success of Mistral Saba is closely related to its specially curated dataset. The model is trained using carefully selected language data from the Middle East and South Asia, enabling it to demonstrate higher accuracy and relevance in handling Arabic-related queries compared to other large general models.

Artificial intelligence is sweeping across the globe at an unprecedented speed, but a significant issue is becoming increasingly prominent: while general-purpose AI is proficient in many skills, it often struggles with specific regional cultures and languages. This is particularly true in culturally and linguistically diverse areas such as the Middle East and South Asia, where users urgently require AI models that understand the 'local language'. Although large general models can handle multiple languages, they often reveal a lack of cultural depth and in-depth regional knowledge when dealing with scenarios rich in regional context.

On February 18th, Shanghai Mita Network Technology Co., Ltd. announced the launch of its latest research mode - 'Shallow Research'. This feature aims to quickly gather information and conduct preliminary analysis using AI technology, helping users obtain valuable research results in a short amount of time.

Recently, Tencent E-signature's official account announced the upcoming launch of an AI contract drafting feature. This innovative function will leverage Tencent's Hunyuan and DeepSeek large model technologies to provide users with a convenient one-click contract generation service.

Recently, Tencent rolled out an important update for its AI assistant 'Tencent Yuanbao,' introducing two advanced models—Hunyuan T1 and DeepSeek-R1—both of which possess deep thinking capabilities. This upgrade allows users to better tackle complex issues, enhancing efficiency in work and study. Users simply need to download and open Tencent Yuanbao to experience these new features for free. The design of the deep thinking model aims to solve complex tasks, uniquely understanding multiple dimensions of problems and potential logical relationships.

Recently, Microsoft launched OmniParser V2.0, a new parsing tool designed to convert user interface (UI) screenshots into structured formats. OmniParser enhances the performance of UI agents based on large language models (LLM), helping users better understand and interact with the information on their screens. The tool's training dataset includes an interactive icon detection dataset, meticulously curated and automatically annotated from popular websites to highlight clickable and actionable areas.