Latest news with #HuggingFace

Mid East Info
20 hours ago
- Business
- Mid East Info
Alibaba Unveils Cutting-Edge AI Coding Model Qwen3-Coder - Middle East Business News and Information
Alibaba has launched Qwen3-Coder, its most advanced agentic AI coding model to date. Designed for high-performance software development, Qwen3-Coder excels in agentic AI coding tasks, from generating new codes and managing complex coding workflows to debugging across entire codebases. Built on a Mixture-of-Experts MoE architecture, this open-sourced model Qwen3-Coder-480B-A35B-Instruct, which has a total of 480 billion parameters but activates 35 billion parameters per token, delivers efficiency without sacrificing performance. The model achieves competitive results against leading state-of-the-art (SOTA) models across key benchmarks in agentic coding, browser use, and tool use. Qwen3-Coder-480B-A35B-Instruct achieves competitive results against leading state-of-the-art (SOTA) models across key benchmarks Additionally, Alibaba is open-sourcing Qwen Code, a powerful command-line interface (CLI) tool that enables developers to delegate engineering tasks to AI using natural language. Optimized with custom prompts and interaction protocols, Qwen Code unlocks the full potential of Qwen3-Coder for real-world agentic programming. The model also supports integration with the Claude Code interface, making it even easier for developers to execute their coding tasks. Trained on an extensive dataset of codes and general text data, Qwen3-Coder is engineered for robust agentic coding. It natively supports a context window of 256K tokens, extendable up to 1 million tokens, enabling it to process vast codebases in a single session. Its superior performance stems not only from scaling across tokens, context length, and synthetic data during pre-training, but also from innovative post-training techniques such as long-horizon reinforcement learning agent RL. This advancement allows the model to solve complex, real-world problems through multi-step interactions with external tools. As a result, Qwen3-Coder achieves SOTA performance among open-source models on SWE-Bench Verified (a benchmark for evaluating AI models' ability to solve real-world software issues), even without test-time or inference scaling. Agentic AI coding is transforming software development by enabling more autonomous, efficient, and accessible programming workflows. With its open-source availability, strong agentic coding capabilities, and seamless compatibility with popular developer tools and interfaces, Qwen3-Coder is positioned as a valuable tool for global developers in software development. The Qwen3-Coder-480B-A35B-Instruct model is now available on Hugging Face and GitHub. Developers can also access the model on Qwen Chat or via cost-effective APIs through Model Studio, Alibaba's generative AI development platform. Qwen-based coding models have already surpassed 20 million downloads globally. Tongyi Lingma, Alibaba Cloud's Qwen-powered coding assistant, will soon be upgraded with Qwen3-Coder's enhanced agentic capabilities. Since its launch in June 2024, Tongyi Lingma's 'AI Programmer' feature—offering code completion, optimization, debugging support, snippet search, and batch unit test generation—has generated over 3 billion lines of code. About Alibaba Cloud: Established in 2009, Alibaba Cloud is the digital technology and intelligence backbone of Alibaba Group. It offers a complete suite of cloud services to customers worldwide, including elastic computing, database, storage, network virtualization services, large-scale computing, security, big data analytics, machine learning and artificial intelligence (AI) services. Alibaba has been named the leading IaaS provider in Asia Pacific by revenue in U.S. dollars since 2018, according to Gartner. It has also maintained its position as one of the world's leading public cloud IaaS service providers since 2018, according to IDC.


The Sun
a day ago
- Business
- The Sun
Alibaba Unveils Cutting-Edge AI Coding Model Qwen3-Coder
HANGZHOU, CHINA - Media OutReach Newswire - 23 July 2025 - Alibaba has launched Qwen3-Coder, its most advanced agentic AI coding model to date. Designed for high-performance software development, Qwen3-Coder excels in agentic AI coding tasks, from generating new codes and managing complex coding workflows to debugging across entire codebases. Built on a Mixture-of-Experts (MoE) architecture, this open-sourced model Qwen3-Coder-480B-A35B-Instruct, which has a total of 480 billion parameters but activates 35 billion parameters per token, delivers efficiency without sacrificing performance. The model achieves competitive results against leading state-of-the-art (SOTA) models across key benchmarks in agentic coding, browser use, and tool use. Additionally, Alibaba is open-sourcing Qwen Code, a powerful command-line interface (CLI) tool that enables developers to delegate engineering tasks to AI using natural language. Optimized with custom prompts and interaction protocols, Qwen Code unlocks the full potential of Qwen3-Coder for real-world agentic programming. The model also supports integration with the Claude Code interface, making it even easier for developers to execute their coding tasks. Trained on an extensive dataset of codes and general text data, Qwen3-Coder is engineered for robust agentic coding. It natively supports a context window of 256K tokens, extendable up to 1 million tokens, enabling it to process vast codebases in a single session. Its superior performance stems not only from scaling across tokens, context length, and synthetic data during pre-training, but also from innovative post-training techniques such as long-horizon reinforcement learning (agent RL). This advancement allows the model to solve complex, real-world problems through multi-step interactions with external tools. As a result, Qwen3-Coder achieves SOTA performance among open-source models on SWE-Bench Verified (a benchmark for evaluating AI models' ability to solve real-world software issues), even without test-time or inference scaling. Agentic AI coding is transforming software development by enabling more autonomous, efficient, and accessible programming workflows. With its open-source availability, strong agentic coding capabilities, and seamless compatibility with popular developer tools and interfaces, Qwen3-Coder is positioned as a valuable tool for global developers in software development. The Qwen3-Coder-480B-A35B-Instruct model is now available on Hugging Face and GitHub. Developers can also access the model on Qwen Chat or via cost-effective APIs through Model Studio, Alibaba's generative AI development platform. Qwen-based coding models have already surpassed 20 million downloads globally. Tongyi Lingma, Alibaba Cloud's Qwen-powered coding assistant, will soon be upgraded with Qwen3-Coder's enhanced agentic capabilities. Since its launch in June 2024, Tongyi Lingma's 'AI Programmer' feature—offering code completion, optimization, debugging support, snippet search, and batch unit test generation—has generated over 3 billion lines of code.


Euronews
3 days ago
- Euronews
Microsoft to make Notre-Dame ‘digital twin' in Europe AI push
Notre-Dame Cathedral took five years to rebuild after the devastating fire, but a digital duplicate would have been much easier, Microsoft's president Brad Smith noted last November. On Monday, he announced that the theory was being put to the test and stated that the Cathedral, built in 1345, would be getting a digital twin. The digital replica of Notre-Dame is being created with the French culture ministry's heritage institute and the start-up Iconem. It combines advanced imaging with artificial intelligence (AI) to produce the digital twin. The aim is to 'preserve permanently in digital form every detail of Notre Dame, ensuring that its structure, story, and symbolism are protected and accessible for generations to come,' Brad Smith, the president of Microsoft, said in a blog post. The digital twin would be donated to the French State and 'can be used by preservationists and be displayed in the future Musée Notre Dame de Paris,' he added. The project will also digitise historic opera sets and millions of artefacts with leading French cultural institutions, such as 1,5000 cinematic model sets from shows at the Opéra National de Paris between 1800 and 1914. Microsoft the aim is the project is to make the collections accessible to as many people as possible, via interactive experiences on its platform The precise budget and completion date for the project are not yet known, but Smith told French media that the operation would take at least a year to complete and will probably cost several million dollars. The project follows a partnership with the Vatican to digitise St. Peter's Basilica. Microsoft also announced on Monday that it would expand its innovation centres in Strasbourg to boost the development of Europe's languages in AI models, especially underrepresented European languages, to make the technology more inclusive. Microsoft's Windows currently includes 90 languages, including the 24 official European languages, as well as Basque, Catalan, Galician and Luxembourgish. The American company aims to develop its languages and enrich its stock of multilingual data by collecting data from sound recordings left on the GitHub platform. Microsoft said that it would then host and make the data 'broadly accessible' on the platform Hugging Face. 'The European project' Microsoft is trying to build its brand in Europe and announced earlier this year a plan called European Digital Commitments. The company said its aim is to expand AI and cloud infrastructure, strengthen digital resilience and data privacy protections, enhance cybersecurity, and support Europe's digital sovereignty and broader economy. However, Europe is also trying to push for its own sovereign cloud infrastructure and is championing its own AI large language models (LLMs) to become less dependent on Big Tech companies. This includes, for example, multilingual text data from GitHub and voice data sets. MOIC and GitHub will partner with Hugging Face, a popular collaboration platform for AI model development, to host and make the data broadly accessible.


Business Upturn
16-07-2025
- Business
- Business Upturn
Sonatype Solutions Now Available in the New AWS Marketplace AI Agents and Tools category
By GlobeNewswire Published on July 16, 2025, 22:00 IST Fulton, Md., July 16, 2025 (GLOBE NEWSWIRE) — Sonatype, the end-to-end software supply chain security company, today announced the availability of its entire product suite, including Nexus Repository Cloud , in the new AI Agents and Tools category in AWS Marketplace. Amazon Web Services (AWS) customers can now easily discover, buy, and deploy Sonatype's industry-leading open source and AI management solutions using their AWS accounts to accelerate AI agent and agentic workflow development. The AI Agents and Tools category in AWS Marketplace serves as a centralized catalog for hundreds of AI solutions from trusted AWS Partners. Among those available is Sonatype's industry leading software composition analysis , proactive malicious code protection, and the world's most trusted binary artifact manager, Nexus Repository, helping organizations accelerate secure software development. 'By offering Sonatype through the AWS Marketplace AI Agents and Tools category, we're providing customers with a streamlined way to access our tools for secure, responsible AI adoption, helping them build and deploy AI agent solutions faster, more efficiently and securely,' said Tyler Warden, SVP of Product at Sonatype. 'Our customers across all industries are already using these capabilities to securely integrate, manage, and govern open source AI models with speed and assurance.' Earlier this year, Sonatype announced its end-to-end AI Software Composition Analysis (AI SCA) capabilities that enable development teams to securely deliver AI-powered solutions without compromising security, compliance, or velocity. Sonatype's end-to-end open source AI solutions provide automated policy management, unmatched AI observability and compliance, and a centralized repository for storing and sharing access to the latest Hugging Face models across the organization. New in AWS Marketplace, Nexus Repository , the leading binary artifact manager, is now available via cloud, providing AWS customers with a centralized repository for storing and managing open source software components and AI models. Joint customers are able to quickly offload infrastructure management with instant scale, high availability, and global access. Through the AWS Marketplace AI Agents and Tools category, AWS customers can significantly accelerate their procurement process to drive AI innovation, reducing the time needed for vendor evaluations and complex negotiations. With centralized purchasing using AWS accounts, customers maintain visibility and control over licensing, payments, and access through AWS. To learn more, visit Sonatype's listing in AWS Marketplace . To learn more about the new AI Agents and Tools category in AWS Marketplace, visit . About Sonatype Sonatype is the software supply chain security company. We provide the world's best end-to-end software supply chain security solution, combining the only proactive protection against malicious open source, the only enterprise grade SBOM management and the leading open source dependency management platform. This empowers enterprises to create and maintain secure, quality, and innovative software at scale. As founders of Nexus Repository and stewards of Maven Central, the world's largest repository of Java open-source software, we are software pioneers and our open source expertise is unmatched. We empower innovation with an unparalleled commitment to build faster, safer software and harness AI and data intelligence to mitigate risk, maximize efficiencies, and drive powerful software development. More than 2,000 organizations, including 70% of the Fortune 100 and 15 million software developers, rely on Sonatype to optimize their software supply chains. To learn more about Sonatype, please visit . Disclaimer: The above press release comes to you under an arrangement with GlobeNewswire. Business Upturn takes no editorial responsibility for the same. Ahmedabad Plane Crash GlobeNewswire provides press release distribution services globally, with substantial operations in North America and Europe.


TechCrunch
16-07-2025
- Business
- TechCrunch
Hugging Face bets on cute robots to bring open-source AI to life
Just five days after opening up orders on its Reachy Mini robots, AI developer platform Hugging Face says it has logged $1 million worth of sales. That's not a bad start for a company that's just recently expanded into robotics, and is largely known for letting developers download open source AI models off the internet. On this episode of Equity, Hugging Face co-founder and chief scientist Thomas Wolf breaks down where he sees the company's robotics going. While other startups tackling robots for the home — such as Figure and 1X — want their robots to offload some of your chores, Hugging Face sees the Reachy Mini as more of a hackable, entertainment device. The device is small enough to fit on your desk, has two misshapen eyes with cameras behind them, as well as microphones, speakers, a head that bobs around, and a pair of antennas for robot ears. It's designed to come with some preset apps, but also lets people build their own apps that can run locally through open source software. Eventually, Hugging Face would like to build out a network of apps for Reachy Mini. Wolf even described the Reachy Mini as 'a bit like an empty iPhone,' hinting at how massive he thinks this market is. The Reachy Mini has gone somewhat viral since launch, largely thanks to its friendly and open design, but also its relatively accessible price point that lets consumers try out an AI-powered robot this year. Wolf says that's a key part to this launch — making something people want to see on their desk everyday, and can actually afford. Wolf describes the Reachy Mini as an entry point for getting consumers to be comfortable with robots in their home, and earning their trust. We got into more of that on the show. I was initially skeptical of the Reachy Mini, but I've started to come around to it. I think Wolf and Hugging Face's vision for the AI device makes seems fun, and seems like a great way for kids or people know to open source software to learn about the space. Wolf also says he's interested to see if people will vibe code apps for their robots, an idea I'm fascinated by and slightly scared of. At other points on the show, Wolf takes listeners inside Hugging Face's acquisition of the french robotics startup, Pollen, and his insistence on developing a robot at a low price point. Wolf also believes that open source AI will play a similar role in robotics as it has in software, and he's trying to position Hugging Face to capture that interest. Later in the episode, we also get into: How Hugging Face plans to leap from software to hardware. Hugging Face's ambitions to one day sell a full-sized humanoid robot. The role of privacy in consumer robotics, and how open-source can address it. Equity is TechCrunch's flagship podcast, produced by Theresa Loconsolo, and posts every Wednesday and Friday. Subscribe to us on Apple Podcasts, Overcast, Spotify and all the casts. You also can follow Equity on X and Threads, at @EquityPod.