Tencent

On November 14th, in the news that yesterday, Mi and Tents officially cooperated, the two sides announced that the "Major" platform under the cloud and the Mi Applied Shop would be connected to smart distribution. It was described that the developer, after completing the creation and commissioning of the AI smart body on the tatterer, could mount a single key to the millimetre application store without additional development. The first page of the millet application store is also synchronised with the "AI Smart Area" where users can search or browse their intelligence directly and use it as if they were downloading App. Last May, the twilight was set to a one-stop AI smart-body creation and distribution platform

On the 14th of November, yesterday, the announcement published the third quarter of 2025, with a total revenue of $192.87 billion, an increase of 151 TP3T. Of this amount, To B ' s business (financial technology and business services) received 58.2 billion yuan, an increase of 101 TP3T over the same year, and became the central engine of growth. According to the financial report, the strong demand for AI by businesses has contributed to the double-digit growth of the telecommunication cloud service and the delivery of technical services at micro-credit shops. This quarter, there has been an increase in strategic input from AI, R & D spending has increased by 281 TP3T to $22.8 billion, a single-season new..

On October 23rd, the Mixer Secretary announced that the first version of the Mixer World Model (WorldMirror) was officially released and opened with new support for multi-views and video input, so that a single card could be deployed and seconds created a 3D world. In July this year, the Queue launched the hybrid world model 1.0, the industry ' s first open-source and traditional CG pipe-compatible tropicable world-generated model, with a lit version that can be deployed in a consumer-level video card. As a former (any-to-any) feedforward 3D reconstruction..

On October 13th, it was announced today that the DeepGEM pathological model developed by the Life Sciences Laboratory in conjunction with Guangzhou Medical University Hospital No. 1 and Guangzhou Institute of Respiratory Health has been massively validated in the lung cancer genetic mutation prediction - only a routine pathological slice image, a genetic mutation of lung cancer prediction of 78%~99% within 1 minute. In other words, it's a new path that does not depend on gene sequencing, but only pathological images can use AI to predict mutation. It'll take a week or two to get the results

Yesterday, the original multi-modular biomode model, "HunyuanImage 3.0" was officially released and opened with parameters up to 80B. It was described as the first open-source industrial-grade multi-modular model, as well as the current open-source model with the strongest effect and the largest amount of parameters, with effect on the head closed-source model of the target industry. Mixtures image 3.0 has been significantly enhanced in semantic understanding, aesthetic sense and reasoning, capable of deciphering the complex semantics of a thousand fonts and producing high-quality images. Unlike the traditional multi-model combination, hybrid image 3.0 uses..

On September 10, the news announced yesterday that the hybrid image model was a new open source of 2.1 in support of original 2K and original Chinese and English input. The Quest was also synchronized with the "PromptEnhancer Text Rewriting Model" and the entry of "Putting a Cute Cat" will automatically complete "Orange Short Cats on a Grid Table with cookies on their paws, water-colored wind" and support for a two-way conversion in Central English, with the words "Starcake with Dream" in Chinese, so that it can also be presented with precision and avoid "expressive ambiguity". Mixed Image Model 2.1 Supporting 1k tokens..

On September 9, the news came out today and published a new AI CLI tool, CodeBuddy Code, and announced that the CodeBuddy IDE international version would be open to public survey without an invitation code and open to all users. As a result, the telecommunication cloud became the first industry-wide AI programming tool manufacturer to support the three types of plugins, IDE and CLI, covering the full range of needs from professional developers to zero base users. It was described as free of charge to users of CodeBuddy's domestic product series, with seamless access to DeepSeek..

On September 6, Tencent announced that the new 2.0 version of the "hybrid game" (Tencent Hybrid Game Visual Generation Platform) was released, with new capabilities such as game graph-generated video, custom model training, and one-key character refinement, and significantly improved the game's 2D graph-generated model capabilities, with the graph-generated video and graph-generated model reaching the SOTA level of the industry in the gaming scenario. SOTA level. This upgrade further solves the pain points of dynamic content generation, style customization, and detail optimization in game art design and distribution, and helps game art designers further improve their efficiency in more scenarios. At the same time of this upgrade, the "Mixed...

A few days ago, the latest member of Tencent's Hunyuan 3D world model series, HunyuanWorld-Voyager, was officially released, claiming to be "the industry's first ultra-long roaming world model that supports native 3D reconstruction". The model focuses on the expansion of AI applications in the field of spatial intelligence, and will provide high-fidelity 3D scene roaming capabilities for virtual reality, physical simulation, game development and other fields. According to the official introduction, Hybrid Voyager breaks through the limitations of traditional video generation in terms of spatial consistency and exploration range, and is capable of generating long-distance...

September 3 news, Tencent Youtu Labs yesterday announced open source intelligent body framework Youtu-Agent. it focuses on "out-of-the-box ready to use", claiming that professional developers and AI enthusiasts can quickly get started, build their own intelligent body applications. Tencent official said, Youtu-Agent has three core advantages: performance: in the authoritative test performance, WebWalkerQA based on DeepSeek-V3.1 accuracy 71.47%, GAIA text subset Pass@1 72.8%. do not rely on expensive...

September 1 news, Tencent mixed yuan announced its international translation model open source, for developers to download and deploy for free. It is understood that this model is named Hunyuan-MT-7B, the total number of references is only 7B, supports 33 languages, 5 kinds of Chinese language / dialect translation, is a lightweight translation model with comprehensive capabilities. According to Tencent Hybrid, in the International Association for Computational Linguistics (ACL) WMT2025 competition that ended at the end of August, Tencent Hybrid's Hunyuan-MT-7B (Entry Name: Shy-hunyuan-MT) took all the...

August 23 news, Tencent Cloud Code Assistant CodeBuddy announced that the domestic version of CodeBuddy IDE is officially open for public beta testing, and the latest DeepSeek-V3.1 model can be used free of charge without an invitation code. 1AI with download address (Beta version): Tencent Cloud Code Assistant CodeBuddy - AI Code Editor Officially, CodeBuddy can truly realize the experience of a person into a whole development team, support...

July 27 news, today, in the 2025 World Artificial Intelligence Conference Tencent Forum, Tencent officially released the hybrid 3D world model 1.0, and fully open source. According to reports, this is the industry's first open source immersive roaming, interactive, simulation world generation model, which brings new possibilities for game development, VR, digital content creation and other fields. In addition to the 3D world model, Tencent Mixed Element also disclosed a series of open source plans, including the end-side hybrid reasoning language model, multimodal comprehension model, and game vision model. Tencent Mixed 3D World Model 1.0

June 27, 2011 - Tencent's hybrid model family today announced the release of its newest member, the hybrid-A13B model, which is the industry's first 13B-level MoE open-source hybrid reasoning model. As a large model based on the Mixed-Mode of Expertise (MoE) architecture, with 80 billion total parameters and 13 billion activation parameters, Mixed-Mode-A13B is claimed to be "comparable to the top open-source models in terms of effectiveness, while dramatically reducing inference latency and computational overhead. Tencent hybrid said that this is undoubtedly good news for individual developers and small and medium-sized enterprises (SMEs), extreme conditions only ...

June 16 news, 1AI learned from Tencent's official WeChat public number that today, Tencent officially launched an algorithm competition. In this competition, Tencent will provide baseline models and technical guidance to help participating students face industrial-grade full-modal data problems (including text, vision, and collaborative ID), and comprehensively utilize a variety of branching technologies such as multimodal comprehension, sequential decision making, and real-time generation to enhance cutting-edge AI systematic practical ability. In order to attract global talents, Tencent will provide a million cash prize pool, in which the champion group can exclusively enjoy the 2 million prize money; breaking into the top ten, you can get Tencent's core business direct off...

June 14, Tencent's official public number announced early this morning that at CVPR2025 (one of the top meetings in the field of computer vision), Tencent announced the open source of the hybrid 3D 2.1 model, which is the first industrial-grade 3D generation model open-sourced by the whole chain. The new model not only optimizes the quality of geometry generation, but also opens up the PBR (Physics Based Rendering) material generation large model, which further improves the texture and light and shadow performance of 3D assets, and says goodbye to the "plastic feeling". According to the official introduction, the new model optimizes detail modeling, resulting in higher mesh accuracy and better topological consistency...

May 28 news, tencent mixed yuan public number today announced open source mixed yuan voice digital human model, just a picture and a piece of audio, you can let the protagonist in the picture naturally speak, sing. The released and open source voice digital human model HunyuanVideo-Avatar, by the Tencent mixed yuan video model (HunyuanVideo) and Tencent Music Tianqin Lab MuseV technology jointly developed to support the head and shoulders, half-body and full-body scenes, as well as multi-style, multi-species and two-person scenes, for video creators to provide a high degree of consistency and dynamism of the Video production...

May 23 news, Tencent's QQ browser today launched the industry's first college entrance examination Agent "AI college entrance examination pass", the AI ability into the whole process of college entrance examination and provide AI tools. According to reports, candidates can quickly access the college entrance examination information through AI, with the support of hybrid large model technology, also supports one key intelligent generation and adjustment of personalized volunteer program to help candidates to volunteer, also supports the punch, stable, and keep three gradients. Before the exam, users can also find "AI problem solving and Q&A" and "AI essay tutoring" in AI Gaokao Tong, which can open up the idea of problem solving and writing. ...

On May 21, Tencent Group Senior Executive Vice President Tang Daosheng delivered a personal speech at the Tencent Cloud AI Industry Application Summit, stating that "every enterprise will become an AI company, and everyone can become a 'super individual'". Tang Daosheng said that in the past year, the availability of generative AI has experienced a leap from "quantitative change" to "qualitative change". The breakthrough in the ability to think deeply about models has accelerated the process of industrial popularization. The industry has also gradually shifted from model pre-training to post-training and application building as the core. Tang Daosheng revealed that through Tencent's various businesses, Tencent Cloud found that its...

May 19 news, Tencent official public number announced today: QQ browsing organ declared officially upgraded to AI browser, and on-line QBot, equipped with Tencent mixed yuan and DeepSeek dual model, currently has AI search, AI browsing, AI learning, AI office, AI writing five functions. 1AI learned from the official announcement that QQ Browser has launched a new AI homepage on both the PC and mobile sides. The browser has also assembled a number of Agents (intelligences), and the first batch of Agents being grayed out include the industry's first college entrance exam A...

According to an April 29 report by Interface News, Tencent has carried out a comprehensive restructuring of its hybrid big model research and development system, mainly focusing on the three core sectors of arithmetic, algorithms, and data, and promoting the continuous development of the hybrid big model by optimizing the deployment of the team and increasing R&D investment. According to reports, Tencent set up two new departments after this restructuring: the Department of Large Language Models and the Department of Multimodal Models, which are responsible for exploring the cutting-edge technologies of large language models and multimodal large models respectively, continuously iterating the basic models, and improving the modeling capabilities. At the same time, it further strengthens the big model data capability and platform base construction, in which the data platform department focuses on big model data all...

April 23 news, Tencent today announced that the hybrid 3D generation model new upgrade, v2.5 new version in the modeling of the fineness of a substantial increase. The hybrid 3D AI creation engine has been fully updated to the v2.5 model base, while the free generation quota has been doubled to 20 times per day. The hybrid 3D generation API has also been launched on Tencent Cloud and is open to enterprises and developers. Compared to its predecessor, the hybrid 3D v2.5 model architecture has been upgraded, with the number of total parameters raised from 1B to 10B, the number of effective facets increased by more than 10 times, and the effective geometric resolution...

March 31, according to the "Penguin Education" public news, recently, Tsinghua University Department of Computer Science AIGE Research Center and Tencent's Sustainable Social Value Division (SSV) Digital Education Lab cooperation of Artificial Intelligence general public welfare courses in the penguin education platform on-line. 1AI learned from the official introduction that the course adopts a dual-teacher classroom model, with senior volunteer lecturers teaching online and local teachers organizing student participation and interaction. As of March 14, 2025, the AI general education courses on the Penguin Teaching Platform cover 6 provinces/municipalities directly under the central government, 24 states, 6 cities, and 6 towns in China.

March 26 news, Tencent Yuanbao early this morning announced the completion of two major model upgrades, at the same time access to Tencent's self-developed deep thinking model "mixed yuan T1" official version and DeepSeek V3-0324 two models. According to reports, Tencent Yuanbao upgraded in code generation, mathematical reasoning, long text processing and other aspects of the ability to further improve. At the same time, the official said, Yuanbao is the first batch of AI applications on line DeepSeek V3-0324, from the model on line to access in just one day. 1AI previously reported that Tencent Yuanbao officials last month (February 19 ...