Zhipu

On January 20, the IQ-GLM-4.7-Flash model was officially released and opened today, January 20th. GLM-4.7-Flash is a hybrid thinking model with a total parameter of 30B and a activated parameter of 3B, claiming to be a homogenous SOTA model, providing a new option for light quantification deployments that takes into account performance and efficiency. As of this date, the GLM-4.7-Flash will replace the GLM-4.5-Flash, go online on the open-think platform BigModel.cn and be available for free call..

On January 14, according to the news, Union Hua was declared today as an open-source new generation image generation model, GMM-Image, based on the roll-out of Atlas 800T A2 equipment and the MindSpore AI framework to complete the full process from data to training, the first SOTA multi-modulate model to complete full training on a national chip. GLM-Image combines image generation with language models using an autonomous and innovative "self-return + proliferation decoder" hybrid structure. 1AI with GLM-Image core highlights as follows..

ON JANUARY 12, ACCORDING TO THE NEWS, THE FIRST PIECE OF THE GLOBAL LARGE MODEL WAS PUT ON THE MARKET ON THE MORNING OF JANUARY 8 AT HK$ 116.20 PER SHARE. ACCORDING TO THE NEW WAVE, IN THE "UNFINISHED DEAL" COLUMN, THE CEO ZHANG PENG RESPONDED THAT AGI HAD NOT BEEN ABLE TO DO SO, BUT HAD BEEN LISTED, SAYING THAT ACHIEVING AGI WAS NOT A VERY SIMPLE THING, NOR WAS IT A VERY SHORT-TERM GOAL. IT COULD BE A MARATHON, A LONG RUN, A VERY LONG DISTANCE. ZHANG PENG SAYS, "WHEN EVERYONE CAN'T HOLD ON, YOU'LL GO TO THE SUPPLY STATION AND GET..

ON JANUARY 9TH, BEIJING JI-SHU ZHU ZHUI ZHUI ZHANG SCIENCE AND TECHNOLOGY CO. LTD. (02513.HK) ( "SWI-SHU") WAS OFFICIALLY LISTED ON THE HONG KONG JOINT EXCHANGE ON JANUARY 8TH. THIS MEANS THAT CHINA IS THE WORLD ' S FIRST LISTED COMPANY WITH ITS CORE BUSINESS WITH THE GENERAL ARTIFICIAL INTELLIGENCE (AGI) BASE MODEL. IN AN INTERVIEW WITH THE FIRST FINANCIAL YEAR, CEO ZHANG PENG SAID THAT IT WOULD BE MEANINGLESS FOR HIM TO LEAVE BEHIND THE BASE MODEL AND CHOOSE TO CALL INSTEAD OF THE AGI TO EXPLORE AND PURSUE COMMERCIAL REALIZATION. IT'S KNOWN THAT THE BIG MODEL..

On December 27, in response to a message from the Ingenuity Spectrum to address the high operational threshold of the command line AI programming tool (e.g. Claude Code, Codex, Gemini, etc.), an AI co-development tool called Z Code was recently launched, providing a unified, friendly visualized desktop, using api key, and sliding to experience multiple Agent programming tools. The core idea of the tool is to deepen the integration of AI intelligence, supporting Claude code, Codex and Gemini..

On 23 December, the CSRC's Department of International Cooperation issued a notice for the filing of “total circulation” of unlisted shares in foreign issues and shares in the country, issued by the Beijing Société Supérieure et Sécurité Internationale. The company intends to issue no more than 43,032,400 shares outside the country and listed on the Hong Kong Joint Exchange. Prior to this, on 19 December, the company disclosed its equity book and announced that it was going to Hong Kong to launch the “First Unit of the Large Global Model”. According to the book, companies earned $5.74 million and $124.5 million in 2022, 2023 and 2024 respectively

I WONDER IF YOU'VE EVER HAD A PROBLEM WITH SOME OF THE VIDEOS YOU'VE RECORDED: THERE'S ALWAYS A FEW GOOD CLIPS IN THE LONG VIDEO, BUT THEY'RE SCATTERED EVERYWHERE. IT TAKES A LITTLE BIT OF TIME TO FIND THE PROGRESS STRIPS, TO REMEMBER THE TIME, TO CUT THEM MANUALLY IN THE EDITING SOFTWARE, AND IT TAKES TIME AND EFFORT. NOW THAT AI CAN READ AND READ THE VIDEO, IS IT POSSIBLE FOR IT TO AUTOMATICALLY PICK OUT THE KEY FOOTAGE FOR ME? THE LOGIC OF THE WHOLE TOOL IS NOT COMPLICATED, AND THE PROCESS IS ROUGHLY THREE-STEP: UPLOAD THE VIDEO FILE, LET AI ANALYZE THE VIDEO CONTENT, FIND THE CLIPS WORTH CUTTING, AND THEN I..

Yesterday, December 10th, AI announced the Official Open Source AutoGLM project, which aims to promote "AI Agent, a mobile phone," as the public domain of the industry. First complete link operation: On October 25, 2024, AutoGLM was considered the first AI Agent in the world with Phone Use capabilities to achieve a stable full operating link on the real machine. Cloudphones and security design: 2025, AutoGLM 2.0 launched, using MobileRL, Comp..

Message dated 9 December, AI Commissioner announced and opened a large multi-modular model for the GLM-4.6V series, including: GM-4.6V (106B-A12B): Basic version of cloud- and high-performance cluster scenarios; GM-4.6V-Flash (9B): Light version for local deployment and low-delayed applications. GLM-4.6V raises the context window for training to 128k tokens and achieves the same parameter size in visual understanding precision as SOTA..

On 9 December, according to the news, the Ingenite Spectrum announced the open source ' s core of AI Agent Model AutoGLM, an AI Agent with the capability of "Phone Use" (cell phone operation) capable of steadily completing complex operating processes such as takeout orders, airline bookings and so forth, which can take dozens of steps. This open source means that hardware manufacturers, mobile phone manufacturers and developers can reproduce an AI assistant in their own devices or systems based on AutoGLM, able to "read" the screen and simulate real people to click, enter, slide, etc. According to IT..

ON THE MORNING OF NOVEMBER 16TH, AT TODAY'S 2025 A.I.+ CONGRESS, PRIVATE LIU, DIRECTOR GENERAL OF THE GESTAPO, STATED, “THE INTELLECTUAL SPECTRUM HAS ITS OWN ATTACHMENT TO OPEN SOURCES, PREFERS OPEN SOURCES AND SHARING, AT THE PRACTICAL STRATEGIC LEVEL, OPEN SOURCES BENEFIT THE ARTIFICIAL INTELLIGENCE INDUSTRY, AI NEEDS THOUSANDS OF PEOPLE IN ALL FIELDS TO PARTICIPATE, AND BASIC MODELS NEED A LOT OF PEOPLE TO STUDY AND PLAY TOGETHER.” “WE FULLY SUPPORT THE OPEN SOURCE, WHICH NOW HAS MORE THAN 40 MODELS.” PRIVATE LAU SAID THAT. OF COURSE, HE ALSO POINTED OUT THAT THE THINK TANK WAS ALSO CONSIDERING HOW TO EXPLORE COMMERCIAL GAINS ON AN OPEN SOURCE BASIS THROUGH..

In the news of 1 October, yesterday, the Gendarmería released the GLM-4.6 flagship text model as the latest version of the GLM series, GLM-4.6 being its strongest code Coding model (up 27% compared to GLM-4.5). According to Reuters, after GLM-4.6 was released, CEO Zhang Peng indicated that there was little chance of achieving full super-artificial intelligence by 2030 (note: Artificial Superintelligence). AI, the industry is talking about when the technology will be super..

September 6, 2011 - Anthropic announced earlier that it will stop selling its Claude service to groups majority-owned by Chinese capital, covering mainland China and companies using it indirectly through overseas registrations or cloud services. Subsequently, Wisdom Spectrum AI announced this evening that in order to help developers make a smooth transition and continue to enjoy top-notch AI coding capabilities, Wisdom Spectrum has officially launched the "Claude API User Special Moving Program". 1AI learned from the official announcement that Wisdom Spectrum is fully compatible with the Claude protocol, and users only need to replace the API UR...

September 2, 2011 - Wisdom Spectrum announced the launch of Claude Code developer exclusive package "GLM Coding Plan" from September 1 - $20~100 per month, applicable to Wisdom Spectrum's latest flagship models GLM-4.5 and GLM-4.5-Air. According to the introduction of the first phase of this package to promote Claude Code, it will be expanded to include more Coding tools in the future. According to the report, the first installment of this package will focus on Claude Code, and will be expanded to include more coding tools in the future. 1AI has learned from the official sources that GLM-4.5 has already accessed Claude Code, Cline...

August 21, 2011 - In October last year, Wisdom Spectrum released AutoGLM, the world's first Phone-Use product. Yesterday, Wisdom Spectrum announced that AutoGLM 2.0 has been upgraded once again, and will soon be online with the function of "Timed Tasks". According to Wisdom Spectrum, AutoGLM 2.0 will elevate the Agent application to a new level: the world's first cell phone Agent, available to everyone; pioneering a new technology paradigm of Agent + cloud phone/cloud computer, which doesn't take over the user's cell phone and computer; breaking through the hardware limitations, operating on any device, in any scenario, and...

August 12 news, Wisdom Spectrum AI today launched the world's 100B level of the best open source visual inference model GLM-4.5V (total parameters 106B, activation parameters 12B), and synchronized in the magic ride community and Hugging Face open source. In addition, the API call price is as low as 2 yuan / M tokens for input and 6 yuan / M tokens for output. 1AI learned from the official introduction that GLM-4.5V is based on the new generation of Wisdom Spectrum's flagship text base model, GLM-4.5-Air, which is a continuation of the GLM-4.5 model, and the GLM-4.5 model, which is a new generation of the GLM-4.5 model, is a new generation of the Wisdom Spectrum.

July 29 news, yesterday, Smart Spectrum released a new generation of flagship model GLM-4.5, specifically for intelligent body applications to build the basic model. Hugging Face and ModelScope platform synchronization open source, model weights follow the MIT License. 1AI with the official introduction of the main points are as follows: GLM-4.5 in the inclusion of reasoning, code, intelligent body of the comprehensive ability to achieve Open source SOTA, in the real code intelligent body of artificial comparison assessment, the actual test of the best in China; the use of hybrid expert (MoE) architecture, including GLM-4.5 ...

Yesterday, Smart Spectrum officially launched and open-sourced a new generation of general-purpose visual language models, GLM-4.1V-Thinking, which is claimed to be "a key leap from perception to cognition for the GLM series of visual models". Specifically, GLM-4.1V-Thinking is a general-purpose inference-based large model that supports multimodal inputs such as images, videos, documents, etc., and is designed for complex cognitive tasks. It introduces the "Chain of Thought Reasoning (CoT Reasoning)" on the basis of the GLM-4V architecture, and adopts the "Reinforcement Learning Strategies for Lesson Sampling (RLCS)", which is a system...

July 2 news, this morning, Smart Spectrum open platform industrial ecological conference held in Shanghai Pudong Zhangjiang Science Hall, open source release of a new generation of universal visual language model GLM-4.1V-Thinking. In the Smart Spectrum open platform industrial ecological conference, the Smart Spectrum announced that the Pudong Venture Capital Group and the Zhangjiang Group of the total amount of 1 billion yuan of strategic investment in the Smart Spectrum, and recently completed the first delivery. At the same time, the three parties also launched a cooperation to jointly build a new AI infrastructure. Wisdom Spectrum today officially released and open-sourced the Visual Language Model GLM-4.1V-Thinking...

April 18 news, on April 16, Beijing Artificial Intelligence Industry Investment Fund announced that on the basis of last year's investment, it will continue to invest an additional 200 million yuan in Z.ai, to support the research and development of Z.ai's open-source model and open-source community ecological construction. According to Beijing Artificial Intelligence Industry Investment Fund, Z.ai is the first AI large model enterprise invested by the fund since its establishment, and it is also the fastest growing enterprise at present. Wisdom Spectrum has a comprehensive accumulation of modeling capabilities including text, inference, voice, image, video, code, etc. In addition, it has a perfect commercialization layout, and has the ability to develop open source models and open source community ecology. In addition, the commercialization layout is perfect, with more than one million scale...

April 15, 2012 - Smart Spectrum published a blog post yesterday (April 14) announcing the launch of a new generation of GLM-4-32B-0414 series models with 32 billion parameters, with results comparable to OpenAI's GPT series and DeepSeek's V3 / R1 series, and with support for very friendly local deployment features. The models are GLM-4-32B-Base-0414, GLM-Z1-32B-0414, GLM-Z1-Rumination-32B-0414, and GLM-Z1-Rumination-32B-0414, and GLM-Z1-Rumination-32B-0414.

April 15 news, April 14, China Securities Regulatory Commission official website shows, "big model six little dragons" one of the Beijing Zhi Spectrum Huazhang Technology Co., Ltd. in the Beijing Securities Regulatory Bureau for the counseling record, by the China International Capital Corporation Limited as a counseling agency, plans to complete the counseling program in October 2025, the assessment and evaluation of the initial public offering and preparation of the application documents. The company plans to complete the counseling program by October 2025, conduct assessment and evaluation, and make preparations for the initial public offering and listing application documents. This means that ChipSmart has become the first "Big Model Six Little Dragons" company to formally start the IPO process, and also the first big model company to formally start the...

March 31, 1AI learned from the official WeChat public number of Smart Spectrum, today Smart Spectrum officially released AutoGLM contemplation at the Zhongguancun Forum. Officially, this new intelligent body not only has the ability to conduct in-depth research (Deep Research), but also can realize the actual operation (Operator), really promote the AI Agent to enter the stage of "think and do". AutoGLM contemplation technology evolution path includes: GLM-4 base model → GLM-Z1 reasoning model → GLM-Z1-Ruminatio...

March 4, 2012 - Big model unicorn Smart Spectrum today released the first model of the 2025 open source year: CogView4, the first open source literate graph model that supports generating Chinese characters. It is reported that the model ranked the first in the DPG-Bench benchmarking in terms of overall score, and reached SOTA among the open source literate graph models. (state-of-the-art) among open source graphical models. The model is also the first image generation model to follow the Apache 2.0 protocol. The GitHub page shows that Wisdom Spectrum has adapted and open-sourced the CogView-4 model for diffuser...