Alibaba

On August 18th, Ali Baba released the AI Music Model Happy Shrimp 1.0 in Chinese called "Happy Shrimp Rice". Users can describe their music needs with emotion, story, memory or use of scenery, and the model completes the words, composes, composes and sings at the same time, producing a full song without having to break the hints into a tune, musical instrument or rhythm label. HappyShrimp uses end-to-end wholesome generation to put language descriptions, lyrics, tunes, emotions, age, voice, tunes, BPM and mixers into the same planning process. Officially, this..

On August 18, according to information received from Ali Baba, Ali thousands of questions will be integrated into Apple intelligence to cover the iOS, iPadOS, macOS and VisionOS of the Chinese market. Users do not need to switch between applications, and can call on apples devices to understand text and image, content generation, etc. This cooperation also includes a second, lower-level path: Apple and Ali will train the Chinese market in a joint, monolingual model. If you take on the current stage of capability access, the joint training model goes to Apple Fonda..

On August 17, according to Ari Baba, the cumulative number of downloads from the model family on various platforms over the past six months has exceeded 3 billion. According to the latest statistics of Hugging Face, the platform was downloaded about 2,045 million times during the same period, up from about 418 million in Google and about 227 million in Meta. More than 460 models have been made public and a large number of community fine-tuning versions have been derived. Hugging Face's open source model report also lists thousands of questions as the leading family in the number of derivative models..

On August 11th, Ali Babala asked App last night about the introduction of a fee-paying scheme for professional users, including two separate fee-paying models for office assistant subscriptions and AI video generation. App shows that the office assistant members have a package of flagships, elites, and seniors. Of these, the monthly price of the flagship package is 128 yuan/month, the annual price is 1,499 yuan/year, which can be increased to 20 times; the monthly price of the elite package is 49 yuan/month, the annual price is paid to 568 yuan/year, which can be increased to 5 times; the advanced package is 19 yuan/..

On August 8, Ali Baba announced yesterday the launch of the country's first one-stop AI Voice Productivity Platform CosyVoice Studio. Officially, the platform is based on the self-researched voice model Qwen-Audio, and contains such functions as voice records CosyFlow, voice smarts CosyAgent and audio content creation tools. Users can search "CosyVoice" at iOS, Android, Mac and Windows app stores, limited..

On the 7th of August, Ali Baba Video Generation Large Model Wan3.0 August 6, opened the public survey, generating a single video of up to 30 seconds and supporting input in document formats for the first time: doc, xls, ppt, pdf, md. A single file or link does not exceed 100 MB, 50 pages, and an upload product introduction to PPT can be used to generate a complete image with a hint. Wan3.0 also supports text, pictures, audio and video as reference, trying to keep roles, props, voices, spatial relationships and styles consistent, and can modify images, dramas and platforms..

On August 4th, Ali Baba released Qwen 3.8 yesterday and simultaneously opened the business-level Agent product QwenWork. Qwen 3.8 API has come online to ask the AI platform for a domestic price of 12 yuan per million Tokens, 36 yuan for output and 1.5 yuan for hidden cache life; Qwen 3.8-Max and 27B versions are scheduled to open their weights next week. Model structure: Qwen3.8-Max total parameter volume is 2.4 trillion. Activate parameter 95..

On July 4, Ali Baba confirmed yesterday that he was strategically integrating three of his corporate AI smart body products. This adaptation is based on the desktop AI smart body tool QoderWork, the core competencies of the AI platform, the Al Goku platform, and the AI implementation engine, MuleRun, which has been set up within Aliyun. The new, integrated AI product will be oriented to the corporate productivity scene, and will be taken over by the CEO and CEO in mid-June this year

On June 23rd, Ali Baba released video generation model HappyHorse 1.1 yesterday. Officially, the new version 1.0 has been updated in terms of dynamic performance, body consistency, command compliance, visual sense and audio. On June 23rd, Ali Baba released video generation model HappyHorse 1.1 yesterday. Officially, the new version 1.0 has been updated in terms of dynamic performance, body consistency, command compliance, visual sense and audio. Generating specifications, HappyHorse..

On June 19, according to news, yesterday, Chairman Cai Chongsing of the Ali Baba Group, at the European Science and Technology Summit VivaTech 2026, described the development strategy of Ali Baba All-Ai. Cai Zhongsin argued in a conversation that the total potential market size of AI (TAM) should target human productivity itself. “AI is creating value units that are equivalent to human intelligence and human productivity. And globally, at least half of the 100 trillion dollars of GDP -- 50 trillion dollars, about human productivity and human intelligence, that's the total of AI..

On May 12th, Ali Baba announced yesterday that he would be in full contact with the treasurer, and that the user would be able to complete the selection of the treasurer's goods and compare them to the next list by talking to him in the App; and that the treasurer would be able to use the function at the entry point at the bottom of the "Show" column. Question and answer: Precise recommendation of goods, supporting parameter comparisons, and one sentence for billing, robbery, return for exchange

On April 21st, Ali ATH announced yesterday that HappyHorse-1.0 would gradually open the API testing through the Ali Yun Proving Platform on April 27th, with the first audience being business-level clients. Ali ATH revealed that Ali Baba has been working in the field of AI video generation for many years, and that the HappyHorse project is led by the ATH Innovation Enterprise, which brings together a number of teams such as the Ali Platform Technology, the General Laboratory and the Nautical Technologies. After opening the test at the end of April, HappyHorse-1.0 will be..

On April 13, Beijing time today afternoon, HappyHorse, who recently lost his red video model, launched the official tweet “@HappyHorse_AI” and released his first article. I'm Happy House, thank you so much for your attention! HappyHorse, an internally measured product of the Ali ATH Innovation Department, is not available, and the " official networks " that are being circulated online are not real. HappyHorse, we're meeting officially

On April 1st, an Ali Baba image generation and editing unified model, Wan2.7-Image, was officially released today. In response to the pains of aesthetic fatigue and colour out of control in the current AI biography, Wan2.7-Image brings with it a more “living sense” of character generation, precision colour control and super-long text rendering. According to the presentation, the model can "give farewell to the AI standard face" and achieve "thousands of faces" to support a brand-new " palette" with precise control colours; and support 3K token 's super-long text rendering to fill a page of A4 paper. Besides

YESTERDAY, ARI BABA CEDO WU MIO SONG SENT AN IN-HOUSE E-MAIL TO ALL THE STAFF OF THE GENERAL INFORMATION LABORATORY, FORMALLY CONFIRMING THAT THE CHIEF OF TECHNICAL SERVICES, LIN JUN-HYUN, HAD LEFT. IN THE LETTER, WU YUI SONG ANNOUNCED THAT THE COMPANY WOULD SET UP A BASIC MODEL SUPPORT GROUP, WHICH WOULD BE CO-ORDINATED BY WU YUI MYUNG HIMSELF, THE HEAD OF THE GENERAL LABORATORY, ZHOU YAN MAN AND FAN XIAN, TO SUPPORT THE BUILDING OF THE BASIC MODEL. APPSO UNDERSTANDS THAT A NEW ROUND OF ORGANIZATIONAL STRATEGIC ADJUSTMENTS IS UNDER WAY WITHIN ALI, WITH A PLAN TO UPGRADE THE BASE MODEL AS A WHOLE, AND TO BRING IN THE TOP TECHNICAL SKILLS ON A LARGE SCALE

On 6 March, according to information received, earlier today, Ali Baba approved the resignation of Lin Jun-hyun of the Tunyi Laboratory. 1AI notes that Omar Sanseviero, head of the DeepMind development team, is on social platform X calling Qwen's friend: “If you want to find a new place to build a good model and contribute to an open model ecosystem, contact us! There are many exciting elements on our road map for development, and much work remains to be done in the future.” On the morning of March 4th, Lin Jun-hyun wrote in social media from..

A message dated 10 February was published for the first time on the GitHub Open Source page of HuggingFace. Qwen3.5 was incorporated into Transformers ' new PR (submission of code consolidation applications). The Information previously reported that Qwen3.5 will open during the Spring Festival. Earlier this month, Tang Jie, the chief intellectual scientist, also revealed on Weibo that many new models, including DeepSeek v4, Qwen 3.5, and GLM-5, would be available in the near future. I don't know

ON 26 NOVEMBER, ALI BABA GROUP PUBLISHED A FINANCIAL PAPER FOR THE SECOND QUARTER OF FISCAL YEAR 2026, GROUP REVENUE OF $24,779.95 BILLION, SUPERMARKET EXPECTATIONS, REMOVAL OF THE IMPACT OF SOLD OPERATIONS, AND INCOME GROWTH OF 151 TP3T PER YEAR. ACCORDING TO ALI BABA CEDO, ARTIFICIAL INTELLIGENCE BUBBLES ARE UNLIKELY TO OCCUR IN THE NEXT THREE YEARS. OVER THE NEXT THREE YEARS, ARTIFICIAL INTELLIGENCE RESOURCES AS A WHOLE WILL BE ON DEMAND. "WE ARE IN THE INPUT PHASE OF BUILDING THE AI TECHNOLOGY AND INFRASTRUCTURE PLATFORM, AS WELL AS THE LARGE-SCALE CONSUMPTION PLATFORM THAT COMBINES LIFE SERVICES WITH ELECTRICITY PROVIDERS TO CREATE LONG-TERM STRATEGIC VALUE. THIS QUARTER..

On November 25th, Ariyun, under the umbrella of the Ali Baba Group, announced on November 24 that technical support would be provided for the latest version of the “Sea-Lion” Large Language Model (LLM), known as Qwen-Sea-Lion-v4, of the Singapore National Artificial Intelligence Programme (AI Singapore, ASG). It's worth mentioning that the AISG this time abandoned the Meta model to the Qwen Open Source Architecture of Ali Baba. In a press release, Aliyun stated that the latest version was based on Aliyun's..

On November 17, Ali Baba today announced that App will be on the line, using the strongest model to build the best personal AI assistant, based on the first open-source model worldwide. According to the official presentation, App is not only able to “talk to” users, but also to “do things” for users and to create future AI life portals, which will cover several life scenarios, including offices, maps, health and shopping. Currently, the app app is installed in major application stores, with web pages and PC versions being made available simultaneously. App countries..

On November 14th, according to Wall Street News, Ali Baba has secretly launched the "Major Question" project to launch a full-scale C-end AI application market against the ChatGPT. According to the report, the Ali Baba Group is preparing a major adaptation of its flagship AI application of the "Tun-Tun-Tun-Tun-Tun " to the name "Mun-Qun" App to locate personal AI assistants and synchronize its development of an international version to compete directly for overseas users. According to the information received, Ali Baba has deployed over 100 engineers at Hangzhou headquarters, dedicated to two-storey office buildings to advance research and development. This layout is considered to have announced 380 billion yuan at the beginning of the year..

Ali Baba ' s Pantyan 2.1 (Wan2.1) is now the leading open-source video model that has become an important tool for AI video creation, based on its strong capacity for generation. This paper will teach you how to quickly install and use the Wan2.1 model to make text-generated videos easy. 1 ‍. Environmental readiness, prior to start, ensure that your equipment meets the following requirements: System support: Windows, MacOS hardware requirements: it is recommended to use equipment with NVIDIA graphic cards, it is recommended to display 16 ≥, it is recommended to use RTX 3060 and above to ensure that models..

On October 10th, at the All-In 2025 Summit, Cai Chongsing, co-founder of the Ali Baba Group and Chairman of the Board of Directors, shared his latest views on hot-spot issues such as Sino-American competition for technology, Ali's strategic transformation and AGI, stating in plain terms: “AI is not a win-winner's area, and the real victory lies not in the development of the strongest AI model, but in the ability to bring technology down and apply more quickly.” Cai Zhongsin and the White House, known as "the tsar of the White House " , AI, and David Sacks, head of encrypted money..

On 25 September, according to the newspaper Kotsing, Ali Yunmong, Ali Baba officially announced his cooperation with Inweida in Physical AI. The collaboration covered various aspects of Physical AI practice, including data synthesis processing, model training, environmental simulation intensive learning and model validation testing. It has been learned that the Ariyun Artificial Intelligence Platform, PAI, will be integrated into the United Kingdom of Great Britain and Northern Ireland Physical AI repository, which will provide pre-processing of data, simulation of data generation, model training assessment, robotics..