Zhipu AI released a major upgrade of AutoGLM at the OpenDay event, enabling it to perform complex operations of more than 54 steps and achieve cross-application task completion. This breakthrough marks significant progress in the ability of intelligent agent technology to complete autonomous tasks. AutoGLM's efficiency even surpasses manual operation, demonstrating its powerful ability and accuracy in handling complex tasks. This upgrade not only improves the technical level of AutoGLM, but also closely integrates AI technology with daily applications.
At today's Zhipu Agent OpenDay event, Zhang Peng, CEO of Zhipu AI, announced the latest upgrade of the company's agent AutoGLM. This upgrade enables AutoGLM to support the execution of complex operations of more than 54 steps and to complete tasks across applications, marking an important step forward for agent technology in autonomous task completion.
Zhang Peng demonstrated the new functions of AutoGLM at the conference. The speed of AutoGLM even exceeded manual operation. This demonstrates the high efficiency and accuracy of AutoGLM in handling complex tasks.
In addition, Zhang Peng also announced that AutoGLM will launch large-scale internal testing and plans to bring it to the market as soon as possible and become a consumer-oriented product. At the same time, AutoGLM also launched the "Free Auto Upgrade for 1 Billion Apps" program, aiming to invite application partners to jointly explore the application potential of AutoGLM in new scenarios.
It is reported that AutoGLM can simulate human operation of mobile phones to perform various tasks, marking a new development of artificial intelligence in the field of "Phone Use". The application of AutoGLM makes AI technology closer to people's daily lives, and can perform a variety of tasks including likes and comments on WeChat, shopping on Taobao, booking hotels on Ctrip, buying train tickets on 12306, ordering takeaways on Meituan, etc. Theoretically, AutoGLM can complete any operation that humans can do on visual electronic devices. The operation logic is similar to humans, and no complex workflow is required.
Test experience address:
1. Qingyan plug-in:
https://new-front.chatglm.cn/webagent/landing/index.html?channel=ads_news_openday
2.autoglm-Android:
https://agent.aminer.cn/
3.glm-pc：
https://www.wjx.top/vm/mOs9cHw.aspx
The release of AutoGLM heralds that AI technology will be further integrated into people's daily lives, bringing users a more convenient and efficient service experience. The launch of this large-scale internal test and the "Free Auto Upgrade for 1 Billion Apps" program also marks AutoGLM's official move towards marketization, and its future development is worth looking forward to.