AI News

Summary February 10, 2025: DeepseekR1 and V3 support on single GPU (24GB RAM) / multiple GPUs and 382GB RAM with up to 3~28x speedup. Hello everyone, KTrans...

In today's rapid development of artificial intelligence technology, voice cloning technology has also ushered in a new breakthrough. The second generation of GPT-SoVITS, jointly developed by the founder of RVC voice changer "Flowers don't cry" and Rcell, the developer of AI tone conversion technology Sovits, has been officially released. This advanced sound...

Artificial Intelligence (AI) systems will profoundly change the way people work in the coming years. To gain a deeper understanding of the long-term impact of AI on the labor market and economy, Anthropic has launched the Anthropic Economic Index (Anthropic ...

Viggle recently launched an innovative feature called "Mic" designed to enhance the user-character interaction experience. Released on October 18, 2024, this feature marks an important advancement in the integration of voice and animation in ViggleAI. Features Advanced...

Preface As the saying goes, "A good tool for a good job is a good tool for a good job". Choosing a good tool is especially important for programmers. Whether it's for automating code or debugging, AI coding assistants can save you time and improve the quality of your code. In this article, we'll explore some of the most frequently used by programmers around the world...

AI is changing the game, and one of the tools that's garnering a lot of attention is DeepSeek - a Chinese version of the ChatGPT alternative.DeepSeek is rapidly emerging globally, attracting a large number of users with its bilingual capabilities and unique features...

Nowadays, the performance of domestic and foreign big models such as DeepSeek is becoming more and more powerful, and the industry generally believes that AI applications will usher in explosive growth in 2025. However, for enterprises, even with powerful big models, they still face the problems of unclear application scenarios and uncertain application forms...

Recently, the Central Office of Internet Information Technology launched a special action called "Clear - Rectify AI Technology Abuse", which draws a clear red line for the governance of a number of issues that have arisen in the current development of artificial intelligence. This initiative aims to guide the healthy development of AI technology and prevent potential risks. The special action focuses on 13 key directions and is divided into two phases...

NVIDIA CEO Jen-Hsun Huang hails AI intelligences as the "digital workforce," and he's not the only tech leader to hold this view. Microsoft CEO Satya Nadella also believes that intelligent body technology will fundamentally change the way businesses operate. These intelligent bodies are able to work with external labor...

Abstract This system card introduces Claude 3.7 Sonnet, a hybrid inference model. We focus on measures to reduce harm through model training and the use of surrounding safety and security systems and assessments. We conduct an extensive review based on the Responsible Extension Policy [1]...

After the verification of the partners, now the free FLUX model generates pictures that are not inferior to DALL-E-3 and MJ. The next step after Controlnet and ipadapter development, you can freely control the picture composition style and so on, DALL-E-3 and M...

Software development is undergoing a profound transformation driven by artificial intelligence (AI). From code completion and bug fixing to generating code through natural language conversations, AI is evolving at a remarkable pace. Recently, ByteDance re-launched a new AI coding...

Google DeepMind released Gemini 2.5, its purportedly smartest family of AI models, on March 25, 2025 (last updated March 26).The first unveiling of Gemini 2.5 Pr...

Cline is an open source VSCode programming plugin, as an alternative to Cursor, which is currently the hottest of the AI IDE tools, and to my surprise the open source software actually offers a free API... Learn more about Cline by reading...

Chinese deep customization Trae based on the Chinese environment for a lot of deep customization, whether it is the interface or interaction, or the content, completely Chinese culture. Use and get started without difficulty Directly from VScode or Cursor to import the configuration This is from ...

Competition in the field of science and technology is always surging. Recently, the Chinese AI startup DeepSeek team updated its V3 base model in a low-key manner without large-scale publicity, and the new version DeepSeek-V3-0324 has been quietly launched on H...

After six months of waiting, it's finally here. After dozens of small version updates, now Topaz Photo AI 3.4.3 version has been optimized in every aspect, and it can be sure that it is more powerful and better to use. This time it is brought exclusively Chinese, after a lot of time research...

The AI.com domain name is really the "meat and potatoes" of the domain name world, and everyone wants it. Think about it, two letters, and with the hottest AI side, is simply a "golden sign". Previously, it was like "choosing a concubine", and a moment later, it jumped to OpenAI's ChatGPT...

Background With the rapid development of cloud computing and artificial intelligence (AI) technologies, online integrated development environments (IDEs) have become an important tool for modern development work. Especially in today's increasingly popular AI and cloud development, online IDEs can not only eliminate the tedious local environment configuration, but also provide powerful...

Comprehensive Introduction Xorbits Inference (or Xinference for short) is a powerful and versatile library focused on providing distributed deployment and services for language models, speech recognition models, and multimodal models. With Xorbits Inference...

Building a local Deepseek AI inference server First the good news! Digital Spaceport is running great performance with the AMD EPYC Rome platform used in previous reviews :😁: This configuration is a real classic! With this...

On January 17, 2025, the Harvard Graduate School of Education (HGSE) released the guide "GenAI in Student-Directed Projects: recommendations and insights," which was developed by the Harvard Creative Computing Lab (C...

Compared to computers, cell phones, which we can't put down, accompany us for a longer period of time and are closer to our lives. If "Computer Use" opens up a new paradigm of human-computer interaction, then "Phone Use" takes it a step further, unlocking more application possibilities and allowing AI to ...

As the AI arms race intensifies, Elon Musk's xAI is once again in the tech world's spotlight with an unexpected move. Shortly after the release of the Grok 3 model, xAI announced that it was making it freely available to the public...

General Description CamON Live Streaming is an application that turns an Android device into a wireless IP camera. With this app, users can repurpose an old smartphone or tablet into a powerful surveillance camera.CamON Li...

Recently, DeepSeek, a Chinese AI startup, launched a new inference model, DeepSeek R1, which has attracted a lot of attention for its outstanding performance. However, a new security assessment has revealed a disturbing fact: DeepSeek R1 ...

In March 2024, a new AI company entered the spotlight with impressive backing: a $21 million Series A led by Founders Fund and backed by a team that included the Collison brothers, Elad Gil ...

In the ever-evolving field of audio production, artificial intelligence is making significant strides, providing a set of tools that can revolutionize the way creators approach sound design. For podcast producers, musicians, and content creators, these advances mean more efficient workflows and higher quality audio effects...

AI innovator LangChain today announced the launch of LangMem SDK, a cutting-edge development toolkit designed to empower AI Agents with long-term memory capabilities. The release of the LangMem SDK marks the first step in the development of Agent technology...

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, it's as if he has 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 A...

There is a paradigm shift in the way humans interact with machines. This is due to the evolution of the conversational Chatbot into an autonomous Agent with hands, brain, and eyes. As one of the first big model companies to explore Agents, Wisdom Spectrum today brings several new developments...

What do you get when you add "MoE" plus "Lightning Attention at an unprecedented scale for production environments" plus "software and engineering refactoring from the framework to the CUDA level"? "What do you get? The answer is a model that equals the capabilities of the top tier models and improves context length...

Today we will delve into the world of open source language models (LLMs) and learn about a new contender for this innovation - Code Llama 70B using OctoAI. Developers often find it difficult to efficiently incorporate a code like Meta's Code Llama 70B...

"360AI search" renamed "nano search", domain name change to https://n.cn/; "360AI assistant" renamed "Nano AI Assistant", the domain name should not have time to change, or the probability of integration into the nano search (looks like it has been integrated). When opening s...

Reasoning is unpredictable, so we have to start with incredible, unpredictable AI systems. Ilya has finally shown up, and right off the bat, he's got something amazing to say. This Friday, Ilya Sutskever, former chief scientist at OpenAI, spoke at the Global ...

What is flowith 2.0? flowith is a next-generation AI authoring platform dedicated to exploring more natural and efficient interactions between humans and AI. In version 1.0, we gained hundreds of thousands of users through canvas-based interaction. In the upcoming...

Just last night, news of Anthropic's upcoming release of a new model spread quickly through the AI community, but not the previously expected Claude 4.0, but rather the Claude 3.7 Sonnet version. Early this morning, Anthropic approved...

Recently, Google Search launched an experimental feature in its generative experience (SGE) called "AI Mode" (AI Mode), which is powered by the Gemini 2.0 Big Model, capable of handling complex queries and with powerful reasoning capabilities, directly...

A Review of DeepSeek-R1's Stability on Third-Party Platforms The rapid growth of the artificial intelligence field has given rise to many remarkable inference models.DeepSeek-R1 has quickly become the center of attention in the industry for its outstanding performance and ability to handle complex tasks. However, with the ...

Anthropic Inc. today released Claude 3.7 Sonnet ^1^, which is not only Anthropic's smartest model to date, but also marks the first hybrid inference model on the market.Claude 3.7 Son...

Today, we are pleased to release FLUX.1 Tools, a modeling suite designed to add control and steerability to our base text-generated image model, FLUX.1, and to support the modification and re-creation of real and generated images. At the time of release, FLUX.1 T...

With the wave of digitization sweeping the world today, data has become the core asset of enterprises. How to quickly extract valuable information from massive business data and present it in an intuitive and easy-to-understand way is the key for enterprises to stand out in the fierce market competition. At the same time, an efficient information transfer mechanism is like the nerve vein of an organization...

Disrupting Traditional Video Generation Kling AI's "Custom Model" feature allows users to train their characters by uploading 10 to 30 videos (each at least 10 seconds long). The process is very different from traditional image training models, as Kling AI utilizes video footage for character...

Since OpenAI's introduction of Function Calling in 2023, the industry has been thinking about how to build a thriving ecosystem of AI intelligences (Agents) and tools to use them. As the underlying models become more robust, the intelligences...

The Computer Use track is crowded with startups and dark horses, as well as giant bigwigs, and now OpenAI is killing it. You can follow Desktop Automation Intelligentsia , we collected dozens of related products. The artificial intelligence field is becoming increasingly competitive, with not only startup newcomers emerging...

Model Overview In recent years, large model training based on Mixture of Experts (MoE) architecture has become an important research direction in the field of artificial intelligence.The Qwen team recently released the Qwen2.5-Max model, which uses over 20 trillion to...

With the rapid development of intelligent body (Agent) technology, MCP (Model Communication Protocol) protocol has gradually become the core communication standard in the intelligent body ecosystem due to its ease of use and efficiency. More and more head AI code editors...

There hasn't been a good voice over product for content production in China, either you can only use the API or the product is okay sound modeling doesn't work. For example, the overseas ElevenLabs, although the English is OK, but the Chinese is really pulling across, the main problem of the open source model is that the model quality is relatively poor...

AI programming tool Cursor has officially released version 0.46, which is a rich update that not only renovates the user interface, but also introduces a new Agent mode and further optimizes Claude 3.7 series....

General Introduction Transformers.js is a JavaScript library provided by Hugging Face designed to run state-of-the-art machine learning models directly in the browser without server support. The library works in conjunction with the Hugging Fa...

Phi-4, a 14 billion parameter open source model with performance comparable to Microsoft's OpenAI GPT-4o-mini, is now available on Ollama. ollama run phi4 main ...

Each of these knowledge points has different content for teachers and students. In 2024, the Massachusetts Institute of Technology (MIT) launched the Day of AI program to fire up the world...

Summary: Elon Musk and xAI have just released the weights and architecture of their massive 314B-parameter language model Grok-1 under the open source Apache 2.0 license. We released our 314 billion parameter Mixtu...

OpenAI recently integrated its advanced image generation technology directly into ChatGPT, a move that quickly ignited user enthusiasm and a chain reaction. The feature leverages the powerful GPT-4o modeling capabilities of the technology lineage with the video generation model Sora...

Recently, Amazon has been making a lot of moves in the field of AI, not only upgrading its consumer voice assistant Alexa, but also forming a new team focused on AI Agent within AWS. This series of moves shows that Amazon is actively laying out the AI Agent field...

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, divided by functional categories, square ...

Artificial intelligence research company Runway recently released its next-generation media generation AI model series, Runway Gen-4, which aims to address the consistency challenges prevalent in current AI video generation and improve the controllability of content generation, marking a move toward a more stable...

Swedish startup Lovable's recent growth trajectory is compelling, demonstrating the potential of AI applications in specific market niches. Founded in 2023 by Anton Osika and Fabian Hedin, the company was initially founded through a program called ...

Snowflake is pleased to announce the release of Arctic Embed L 2.0 and Arctic Embed M 2.0, the next iteration of our cutting-edge embedding model, which now supports multilingual search. While our previous releases suffered from...

Dear Developers: We are excited to announce that Devv is building a disruptive product, Devv Builder, which is about to be officially released this month! Devv Builder is a revolutionary Coding Agent specifically designed to generate...

Github: https://github.com/hkust-nlp/simpleRL-reason This blog will show a replication of DeepSeek-R1-Zero and DeepSeek-R1 training...

The water droplets attacked the solar system using the most primitive form of attack: impact! And yet it's so efficient. After we've messed around with containers, orchestration, CI/CD, DevOps, and invented a bunch of terms that don't do much good, we eventually realize that all of these operations are just fancy and get developers deeper and deeper in the hole. Eventually...

After being deeply involved in AI-assisted development for the past few years, I've noticed an interesting phenomenon. While engineers report significant productivity gains from using AI, the actual software we use on a daily basis doesn't seem to be significantly better. What's going on here? I think I know why, and the answer reveals that we...

The Dify team is excited to announce a major v1.0.0 release update for Dify, the AI application development platform! This milestone release marks a solid step forward in building Dify's next-generation AI app development platform, centered around a new plugin...

In today's rapidly changing technology, artificial intelligence (AI), as a force to be reckoned with, is profoundly changing the way we live and work. And in the surging wave of AI, Ten Square Ronghai Xiaozhi AI Voice Box has jumped to the top of GitHub with its excellent performance and innovative user experience...

The benchmark evaluation of Chinese big model "AI Search" (SuperCLUE-AISearch) was released, which is an in-depth evaluation of the ability of big model combined with search. The evaluation not only focuses on the basic capabilities of the big model, but also focuses on its performance in scenario applications. The assessment covers five basic capabilities such as...

SYSTEM DESIGN PRINCIPLES The optimization goals of the DeepSeek-V3/R1 reasoning service are: higher throughput and lower latency. To optimize these two goals, DeepSeek employs a solution called cross-node expert parallelism (EP). First, EP significantly scales the batch ...

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

The long-awaited Trae Windows desktop version is officially open for download today (February 17th)! Click to download: Trae-Setup-x64 Trae for Windows version 100% replicates the macOS interface, operating...

Everyone is using AI tools, we watched AI develop and grow step by step, most of the time we used to just use text to chat with them, and there were times when Kernel wondered: it would be nice if we could think well about pictures sometime. After researching AI a bunch of times, we used Kimi later on and realized that it's...

In the current wave of rapid development of AI technology, several open source projects provide developers with powerful features and flexibility, especially in the field of Multi-Agent Systems (Multi-Agent Systems). Here are five of the most popular Agent projects (GitHub...

The development of AI models is becoming more and more diversified. In addition to large-scale language models and small-scale language models, "world models", which are called world simulators, are being regarded as one of the next key development directions of AI. 2...

Copilot for PowerPoint is very powerful to create PPTs directly from natural language, PPTs from other files (Word, PDF or PPT): At Ignite 2024, Microsoft released the new Copilo...

Google has come up with a big wave again, yesterday shared the Gemini 2.0 Flash native image generation and editing capabilities, today Gemini has been paid to use the Deep Research tool is now free to use. Many people still don't know...

Domestic AI company DeepSeek recently released its latest model DeepSeek V3, which is said to be one of the strongest open-source AI models, similar to the famous ChatGPT, Claude and these models. Now the web version of DeepSee...

Google recently announced that its video generation model, Veo 2, has been officially integrated into the Gemini Advanced service as well as its experimental platform, Whisk. The move means that Google One AI Premium subscribers can now...

Introduction In the rapid development of artificial intelligence technology today, all kinds of AI tools have penetrated into every aspect of our life and work, from audio processing, document interaction to content creation, image generation, and then to software development and data analysis, AI is changing the traditional workflow and industry pattern at an alarming rate...

For well-known reasons, it is not possible to access ChatGPT normally in the country, OPENAI does not provide services directly to the country, and the country has blocked the official website of OPENAI, so there are all kinds of ChatGPT pay to open services in the network. First of all, most users in the domestic network environment...

Heavy news in the AI circle exploded. Alec Radford, the legendary OpenAI researcher known in the industry as the "father of GPT", announced that he was leaving the company to pursue independent research. As the chief designer of the GPT series, the core technology behind ChatGPT, Radford's decision...

Do not access ONE-API and apply for multiple accounts as a staging service load! One is unethical and the other is blocked. MistralAI launched the API free package has been a few days, has been hesitant to send out the reason is that Mistral has been repeatedly fooled developers.... ...end of August...

Making predictions, especially in a fast-moving field like data and AI, is notoriously difficult. Nonetheless, we, Rajesh Parikh and Sanjeev Mohan, released our 2024 trend predictions last year. With 2024 ...

Comprehensive Introduction Laminar is an open source AI engineering optimization platform focused on AI engineering from first principles. It helps users collect, understand, and use data to improve the quality of LLM (Large Language Model) applications.Laminar provides comprehensive observability, text...

Introduction DashInfer-VLM is an inference architecture for visual multimodal large model VLMs, especially optimized for inference acceleration of the Qwen VL model. The biggest difference between DashInfer-VLM and other inference acceleration frameworks for VLMs is that it puts the VIT part of...

In 2021, Microsoft launched GitHub Copilot, which jumped to become the most sought-after AI tool in the programming world. GitHub Copilot can automatically generate complete code functions based on contextual information provided by the user, such as function names, comments, code snippets, etc., and is known as a coding...

Write code at the speed of the Tab key. Available today in Continue.dev and coming soon to other leading AI code assistants. Of all the innovations in AI over the past year, code generation has arguably been the most important. Similar to assembly lines streamlining manufacturing and...

Tech giant OpenAI recently announced a series of major updates, with CEO Sam Altman taking to social media platform X to reveal the latest developments in GPT-4.5 (codenamed "Orion") and GPT-5, as well as ...

The Google App on Apple phones no longer comes with the Gemini feature, so those who want to use Gemini will have to download a separate app to do so. Recently, Google has announced that they are bringing the iOS version of Google ...

On November 20, 2024, OpenAI launched a free online course designed to help K-12 teachers introduce ChatGPT into their teaching. The course is called ChatGPT Founda...

Still stupidly refreshing the queue in the official DeepSeek? Or deploying and stepping on various clouds? Recently, Tencent's intelligent workbench ima.copilot (ima for short) has made an important upgrade at the technical level, quietly accessing the DeepSee...

Recently, an exhaustive list of AI tools has been circulating in the industry, covering a wide range of innovative AI-driven applications at home and abroad. These tools not only demonstrate the wide range of applications of AI technology in various fields, but also signal the arrival of an AI-led productivity revolution. Domestic AI ...

On April 1, OpenAI CEO Sam Altman announced via social media that the company plans to release a new "open-weight" language model with powerful reasoning capabilities in the coming months....

A strategic turn for a star company: from Arc to Dia The Browser Company has been in the spotlight since its inception in New York in 2019. With a cumulative $128 million in funding and a valuation of over $550 million, the company...

Leaky pipes are an undetectable enemy in the battle to conserve water, causing an average of 301 TP3T of piped drinking water to be lost globally, and in some regions this is as high as 701 TP3T. In some regions, the rate is as high as 701 TP3 T. These leaks are often difficult to pinpoint, but an innovative AI technology is changing that...

From Sundar Pichai, CEO of Google and Alphabet: Information is at the heart of human progress. That's why we've been working for 26 years to organize the world's information and make it accessible and useful. It's also why I...

May 5, 2025 - OpenAI, a leader in artificial intelligence, has confirmed a notable deal: the acquisition of AI coding tool startup Windsurf for approximately $3 billion. Not only is this OpenAI's biggest deal to date...

Day of AI (Artificial Intelligence Day) was born in June 2021 at MIT. Led by a team of faculty and research scientists from MIT's RAISE (Responsible Artificial Intelligence and Education), the Day of AI program aims to do for students of all backgrounds and abilities...

Early last year, Zhu Xiaohu was interviewed by Zhang Xiao-jun. At that time, the Dark Side of the Moon had just completed a $1 billion financing round, and the user growth figures were skyrocketing; OpenAI released the technical report and demo video of Sora, and the entire AI circle was once again in a frenzy. The most popular topic at that time was the...

Windsurf is releasing a preview version called Windsurf Next, which is intended for users who want to get a taste of the latest features, even if they are not quite perfect and may still have some minor issues that need to be addressed in the official W...

Model Context Protocol (MCP) is becoming a hot topic in the circles of building AI applications and agents. Much of the discussion centers around installing and running an MCP server on a local computer...