AI News

Mistral AI has recently announced the release of its latest model, Mistral Small 3.1, which it claims is the best of its class today. This new model builds on the foundation of Mistral Small 3, with significant improvements in text performance, multimodal understanding, and contextual processing capabilities,...

In the era of information explosion, how to quickly and accurately locate key information from massive data has become the core challenge of enterprise and personal knowledge management. Recently, the Dify product team released v1.1.0 and innovatively introduced the "metadata" as the core of the knowledge filter function. This update is like...

OCR technology is capable of converting textual information in an image into editable and processable text data. Simply put, it recognizes and extracts text from images. Next, we will review the 10 OCR open source projects with the highest number of Stars on GitHub to provide you with a detailed guide to choosing an OCR tool. 01 GOT-OCR 2.0: end-to-end multimodal OCR model GOT-OCR 2.

  Gemini has been updated a bit frequently lately, in no particular order: Veo2 inference model is now live in Google AI Studio and Gemini (shrunken version) Native support for multimodal models for image generation and editing: Gemini 2.0 Flash (now standardized as: Gemini 2.0 Fl...

Chinese internet giant Alibaba is making a big push into artificial intelligence (AI). Alibaba CEO Wu Yongming has reportedly made it clear that he wants to fully realize AI-driven in the company's existing businesses. In an announcement on the Hong Kong Stock Exchange (Feb. 24), Alibaba plans to invest at least $380 billion over the next three...

Core Points: The MCP protocol lays the groundwork for a broader range of future applications by introducing a "streaming HTTP" transport scheme that enables complete statelessness and simplifies communication. The recent adoption of a key technical enhancement to the Message Channel Protocol (MCP) signals that this emerging protocol will...

Recently, the emergence of a series of open-source AI Agent (Intelligent Body) frameworks has attracted a lot of attention in the industry. These frameworks are not simple replacements for LangChain, Crew AI, or the OpenAI Agents SDK, but offer unique features and perspectives designed to simplify and accelerate Multi-Agent...

In the field of artificial intelligence, large-scale language modeling (LLM) technology is rapidly changing, and various tool libraries are emerging. In order to help developers better cope with the challenges of LLM development, this paper organizes a toolbox containing more than 120 useful LLM libraries, and divides them by functional categories, so that engineers can quickly...

In the wave of digital transformation, automated workflow tools have become the key to improve efficiency and reduce costs. In the increasingly mature AI technology today, how to combine AI and automated workflow has become the focus of attention in the industry. In this article, we will review three popular tools: n8n, Coze...

According to internal sources, Anthropic is actively working on two new features called Harmony and Compass that are designed to significantly enhance the capabilities of its AI model Claude. These new features are expected to be integrated into Claude to provide users with more powerful code assistance and deep research support. Harmo...

Recently, Google introduced a new experimental text embedding model gemini-embedding-exp-03-07[1] in the Gemini API. The model is trained based on the Gemini model, inheriting Gemini's deep understanding of language and subtle contexts, and is applicable to a wide range of scenarios. It is worth mentioning that this ...

Google has announced an experimental feature for its Gemini AI assistant called Gemini with personalization. This new feature will allow Gemini to connect to a user's Google apps (currently supporting Google Search History first) to provide more...

On March 16th, Baidu officially released two new big models: Wenshin Big Model 4.5 and Wenshin Big Model X1, which are already online on Wenshin Yiyan website and users can experience them for free. At the same time, Wenshin Big Model 4.5 is now available on the Baidu Intelligent Cloud Qianfan Big Model platform, where enterprise users and developers...

Sakana AI recently announced that papers generated by its "AI Scientist" system had passed peer review in a workshop at ICLR, the top conference in machine learning. This has sparked a wide-ranging discussion about whether AIs are capable of scientific research. In-depth exploration of the significance of this event...

In the field of Artificial Intelligence coding, how to make AI Intelligent Bodies (Agents) more effectively utilize tools to complete complex software development tasks has been a core issue of great concern. "Tool Use/Function calling" is a key technology born in this context. A well-developed software development ...

Yesterday, Google shared the Gemini 2.0 Flash native image generation and editing capabilities, and today, the Deep Research tool in Gemini, which has been paid for, is now free to use. Many people still don't know what Deep Research is, or have heard a lot of introductions...

Have you ever had the experience of working side-by-side with a talented assistant who always understands your needs quickly and gives subtle answers, but after every short break, he seems to have amnesia and needs you to re-explain the project background, technical architecture, and even the most basic requirements from scratch? For those who rely on...

This week, Agent (intelligent body) technology swept through the tech world at an unprecedented rate, and behind this boom is a leap forward in reasoning modeling capabilities. On the evening of March 5, Manus made a stunning debut, and its powerful demo instantly set off the entire Internet. Just two days later, the domestic team DeepWisdom MetaGPT ...