Gemini 3 late night raid! Powering through GPT-5.1, Google's AI throne is finally secure
Google quietly went live with the Gemini 3 Pro large model at 3am without a launch event. The model topped LMArena with 1501 points Elo, the most human...
Veo 3.1 vs Sora2: Who is the real king of video generation?
Google's Veo3.1 competes with OpenAI's Sora2 in the field of AI video generation.Veo3.1 is known for its precise control, high-quality audio and video synchronization, and is suitable for...
20-year-old college student's coursework, 1 day wildly 4000 + Star, the public opinion analysis rolled on the GitHub hot list of first
BettaFish (Micro Opinion) is an open source AI opinion analysis project developed by 20-year-old college students, originated from a course assignment, and gained 4000+ Star...
NextStep-1: The "Ultimate Form" of Autoregressive Image Generation, 14B Parametric Model Open Source!
The StepFun team has open-sourced NextStep-1, a 14B-parameter pure autoregressive image generation model. The model is directly in continuous visual space...
LTX-2 blew up! The world's first 4K video generation model with synchronized audio and video, supported by ComfyUI!
LTX-2 is the world's first audio-video synchronized 4K video generation model released by Lightricks, which generates 20-second, 50fps HD video with text/image...
Humans Can See, AI Can't: The Essential Difference That Hidden Heart Reveals
A static black-and-white noise map will reveal a dynamic heart pattern through mobile viewing or page zoom, while AI models such as Gemini 2.5 Pro, GPT-5, and Beanbag are...
FlowithOS is online! The world's first "AI Intelligent Body Operating System", can it really be your digital employee?
FlowithOS is the world's first native operating system designed for AI intelligences, capable of performing complex tasks rather than being limited to chat. The system is based on the Chromium browsing...
Cursor 2.0 blew up! Self-developed model Composer debuts, code generation is ridiculously fast!
Cursor 2.0 is officially released with the launch of its own large model Composer. its code generation speed is up to 250 tokens/second, which is faster than GPT-5 and ...
ChatGPT Atlas: a revolution in AI browsers
OpenAI releases ChatGPT Atlas, the first AI-native browser, deeply integrating ChatGPT's intelligent capabilities. Its core features include: real-time AI co...
Blockchain, Bitcoin, and Web3: What's the Relationship Between the Three and Are They Okay in 2025?
Blockchain, Bitcoin, and Web3 have made it clear in 2025 that the price of "digital gold" has surpassed $110,000, reaching an all-time high of $111,013; blockchain has become the "new infrastructure...
DeepSeek-OCR: Opening a New Era of Visual Compression
DeepSeek-OCR is based on a visual language model and "contextual optical compression" technology, which is realized by transforming images into visual tokens...
Manus and the AI Agent Bubble: From Ideal to Disillusionment
Manus, as a representative of the AI Agent boom in 2025, relies on big models, toolchain and memory technology to realize task execution, but due to the lack of professional scenarios of deep cultivation and closed...
Kimi K2 Thinking Suddenly Released! 1 Trillion Parameters Open Source Beast Beyond GPT-5
Dark Side of the Moon releases Kimi K2 Thinking, an open source thinking Agent model with 1 trillion parameters. Its core breakthrough lies in the fact that without human intervention it can continuously...
An article to read about Web3 technologies and applications
Web3 has moved from concept to reality, with a global market size of $21.35 billion in 2025 and a related industry in China of more than $20 billion. Its core lies in user sovereignty, through...
MCP, RAG, Agent, Function Call details and other related AI terms
This paper systematically introduces four core technologies in AI development:MCP (Model Context Protocol) as a standard communication specificationRAG (Retrieval Augmentation Generation) to realize real-time knowledge supplementationA...
KAT-Coder: A New Breakthrough in Racer AI Programming
Racer launched AI programming product matrix KAT-Coder, covering self-developed models, tools and platforms, supporting more than 20 programming languages and multiple types of development tasks. Its open source version KA...
Bolt.new: A Tutorial Guide to Creating Professional Websites with Simple Descriptions
Bolt.new is an AI-driven development platform where users write code by generating full websites directly from natural descriptions. It supports multiple frameworks for generating applications, installing packages...
30 seconds to deploy, let the whole network hotspot actively find you, this magic tool completely solved my information anxiety
TrendRadar is an open source hotspot aggregation tool, by automatically crawling the real-time content of 11 mainstream platforms such as Zhihu, Weibo, Shake, etc., according to the user's preset keywords fine...
Grok 4.1 Low-Profile Release! General Purpose Capability Crushes All, Emotional Intelligence Tops #1
Musk's xAI company lowers Grok 4.1, tops LMArena charts with 1,483 points and EQ-Bench3 emotional intelligence test...
Minute-level real-time video generation is here! Tencent and Nanyang Technological University jointly break the bottleneck of long video generation
The Rolling Forcing method, jointly developed by Polytechnic University and Tencent ARC Lab, solves the difficulty of combining quality, consistency and real-time of AI long video generation...
