Ant Group

On February 11, an ants group open source released a large full-mode model Ming-Flash-Omni 2.0. In a number of open benchmarking tests, the model has been prominent in key competencies such as visual language understanding, voice-controllable generation, image generation and editing. Ming-Flash-Omni 2.0 was described as the first industry-wide unified audio generation model to generate both voice, environmental sound and music in the same track. Users can exercise precision control over sound, speed, tone, volume, emotions and dialects by using only natural language instructions. Model..

On December 16, the AQ brand of the AHRA AQ under the ants Group announced today that it would be upgraded to "Ant Ahf" and released a new version of the App to upgrade the three functions of health companionship, health questions and answers, and health services. The ants officially report that the current monthly users of the "Ant Afu" App are over 15 million, making it the first major health management AI App. It is understood that Afou App is now responding to more than 5 million health questions from users per day and that 55% users are from three and below. The App has been with 8 provincial and municipal health committees, 18 provinces..

ON DECEMBER 9TH, THE ANTS GROUP ' S ALL-MODIFIED GENERIC AI ASSISTANT LIGHT OFFICIALLY LAUNCHED A WEB VERSION, WHICH COMPLETES THE MULTI-END ECOLOGY OF THE LIGHT. USERS CAN ACCESS IT DIRECTLY THROUGH BROWSERS TO EXPERIENCE THEIR CORE `LIGHT DIALOGUE' AND `LIGHT FLASH APPLICATIONS' AT THE COMPUTER END. THE WEB VERSION CONTINUES THE CENTRAL ADVANTAGE OF "30 SECONDS TO GENERATE SMALL APPLICATIONS IN NATURAL LANGUAGES" AND SYNCHRONIZES WITH DATA AND CREATION AT THE MOBILE END. USERS CAN DESCRIBE DEMAND ON THE COMPUTER, QUICKLY GENERATE, EDIT AND MANAGE DIALOGUE AND FLASH APPLICATIONS, AND CONTINUE TO USE AND SHARE THEM AT THE END OF A MOBILE PHONE. THE LIGHT PAGES SHOW THAT THE LIGHT PAGES ARE MAINLY USED FOR WORK AND STUDY..

On September 30th, an ant group announced this early morning the first mega-model of the hundreds of billions of parameters it studied itself, Ring-1T-preview, a large-scale model of natural language reasoning and the first large-scale model of the hundreds of billion parameters of open source reasoning around the globe. Ring-1T-preview is a preview of a mega-mode Ring-1T of trillion parametric reasoning, according to the official information of the Magna Carta Model. Although it was a preview, its natural language reasoning was excellent. In the AIME 25 test, Ring-1T-preview received 92.6 minutes..

August 18th, Ant Group announced today through its official public number that it will declare war on fake medical advertisements. To this end, AQ, Ant's AI health manager, will also officially launch a series of AI initiatives to combat counterfeiting. The program includes four major initiatives.1 AI summary is as follows: Identify fake ads: Input "fake" in AQ App, you can quickly call a number of AI fake functions. These include AI photo fighting, AI phone checking, and 24-hour rolling rumor debunking. Avoiding fake science and technology: launch a special area for combating fake rumors to crack down on health rumors. Fight against fake "miracle doctors": Including joint...

July 28th news, ant group CEO Han Xinyi in 2025 world artificial intelligence conference (WAIC), shared the practice and thinking in the field of AI. Han Xinyi said, science and technology universal benefit, is the ant group unchanging heart and color, in the AGI era, the ant will focus on AI application, so that the breakthrough of the intelligent cap, benefit the daily life of the people. Han Xinyi shared the ant group focus on AI application of three points of thought: AI driven intelligent service is professional, personalized and automated, at the same time, the application of the AI era is not only limited to the digital world...

July 15 news, according to ant technology news, the world digital academy of sciences (WDTA) in the United Nations headquarters in Geneva recently formally released the AI STR series of new standards "AI intelligent body operation security test standard", the standard by the ant group, Tsinghua University, China telecom lead, joint PricewaterhouseCoopers, Nanyang Technological University in Singapore, the United States, Washington University in St. Louis and other twenty domestic and foreign institutions, Enterprises and colleges and universities to compile, for the world's first single intelligent body operation safety test standards. According to reports, the standard addresses the risk of "behavior" brought about by intelligent bodies crossing the "language wall", and for the first time, the input...

June 26 news, Ant Group today released AI health application "AQ", around the people's immediate needs for medical treatment, health management pain points, AQ provides health science, consultation, report interpretation, health records and other hundreds of AI functions, but also efficiently connects the country's more than 5,000 hospitals, nearly a million doctors, nearly 200 famous doctors AI bilocation and other professional medical services.1AI noted that at present the App has been successively launched in major application stores. 1AI noted that the App has been launched in major app stores. Ant Group Vice President Zhang Junjie introduced, in September last year in Alipay on-line "AI health manager", has served...

On May 28th, the Ant-Ling Big Model (Ling) team officially open-sourced the unified multimodal big model Ming-lite-omni. According to the introduction, Ming-lite-omni is an all-modal model based on the MoE architecture constructed by Ling-lite, with a total parameter of 22B and an activation parameter of 3B, which supports "cross-modal fusion and unification", "understanding and unification", and "generation unification". It supports "cross-modal fusion and unification" and "comprehension and generation unification". In several comprehension and generation evaluations, Ming-lite-omni with only 3B of parameter activation, performs as well as 10B of...

March 24 (Bloomberg) -- Ant Group is using Chinese-made chips to develop AI model training technology, which will reduce costs by 20%, according to Bloomberg today. Ant Group is using domestic chips including Alibaba Group Holding Ltd. and Huawei Technologies Co. to develop AI model training technology using a mixed-expert model (note: MoE, Mixture of experts) machine learning method. The report also mentioned that the training results rival NVIDIA's H800 chip. One of the people familiar with the matter revealed that Ant Group is still using the NVIDIA chip...

According to a report by Blue Whale News on November 20, model track startup Biancai Technology confirmed that it has been acquired by Ant Group, and the company will operate independently, but the investor has already withdrawn. The report mentioned that there is news that Ant Group has recently set up a reinforcement learning laboratory, and the founder of the border technology, Wu Yi, has been employed as the chief scientist of the laboratory. Wu Yi responded in a circle of friends on the event of joining the ants, saying: "Recently invited by the ants to do ants big model reinforcement learning direction of the research, my teaching in the Tsinghua University Institute of cross-information, as well as the edge of the science and technology are continuing. Ant has a very good big model application scene, I hope to make an impact together...

At the Bund Conference Wealth Forum on September 6, a new version of Ant Group's AI financial butler "Ant Xiaozai" was upgraded and unveiled in the Alipay App. The upgraded "Ant Xiao Cai" can interpret hotspots in real time and provide "question-and-answer, personalized" services. At the same time, Ant Wealth's new App equipped with "Ant Xiao Cai" Pro version is on line for grey test, bringing a new service experience such as "language control everything, customized briefings", which is also the first financial management App in China to realize AI native experience. According to Ant Group, AI is the most popular way to manage financial affairs. The AI Native Experience is the first AI-native financial management app in China.

Ant Group, together with Beijing News and Beike Finance, has jointly developed and launched a generative big model product called "Beike Finance·Smart ESG". This product aims to provide users with more accurate, comprehensive and valuable analysis and interpretation of ESG (environmental, social and corporate governance) reports of A-share listed companies. It can answer various questions of users in real time and provide in-depth analysis and summary. "Beike Finance·Smart ESG" uses the big model and multi-agent technical support provided by Ant Group's AI industry assistant Zhi Xiaozhu. Based on the big model customized by Ant for rigorous applications, Zhi Xiaozhu combined agentU...

Ant Group open-sourced a new project called EchoMimic on the 10th, which can help people "lip-sync" through facial features and audio, and combine facial landmarks and audio content to generate relatively stable and natural videos. The project has high stability and naturalness. By integrating the features of audio and facial landmarks (key facial features and structures, usually located in the eyes, nose, mouth, etc.), it can generate videos that are more in line with real facial movements and expression changes. It supports the use of audio or facial landmarks alone to generate portrait videos, and also supports combining audio and portrait photos to make "lip-syncing...

The 27th United Nations Science and Technology Conference was held in Geneva, Switzerland. The World Digital Technology Academy (WDTA) released a series of breakthrough results, including the two international standards "Generative Artificial Intelligence Application Security Testing Standard" and "Large Language Model Security Testing Method". It is reported that these two international standards were jointly compiled by experts and scholars from dozens of units including OpenAI, Ant Group, iFlytek, Google, Microsoft, NVIDIA, Baidu, Tencent, etc. Among them, the "Large Language Model Security Testing Method" was compiled by Ant Group. The standards released this time bring together the wisdom of experts in the field of global AI security,…

Ant Group's self-developed intelligent R&D platform CodeFuse has launched a new feature called "Graphics to Code", which supports developers to generate code with one click using product design drawings, greatly improving the development efficiency of front-end pages. The relevant function is currently being tested internally. It is reported that Ant Group is fully promoting AI programming internally, and the number of engineers using CodeFuse to support daily R&D work has reached more than 50%, and 10% of the code submitted by these engineers is generated by AI. Currently, within Ant Group, more than 50% of programmers use CodeFuse in daily R&D every week. Co…

With the rapid development of the era of artificial intelligence, Ant Group has actively increased its investment in the field of AI. It is reported that Ant Group has set up an AI innovation R&D and application department, NextEvo, and appointed Xu Peng as the head of the department. Xu Peng worked at Google for 11 years, during which time he was responsible for and led the R&D of Google Translate's core technology and participated in the development of Google's display advertising system algorithm. It is reported that the NextEvo department will be fully responsible for the core technology R&D of Ant AI, including the R&D of the Ant Lark model. This move aims to further…