Chinese AI start-up Zhipu AI unveiled a free AI agent on Monday, in the latest example of what has been described by some industry leaders as "explosive growth" in China's artificial intelligence industry marked by a rising number of products and technological advancements.
It has been claimed that AutoGLM, marks the world's first AI agent capable of both deep research and practical operations, the company said in a press release sent to the Global Times on the same day.
Specifically, AutoGLM is capable of performing complex "thinking" and executing tasks simultaneously, Zhipu AI CEO Zhang Peng (right) said and adding that it is capable of opening and browsing web pages like a human being and conducting tasks such as data retrieval, analysis and research report writing.
According to Zhang, the AI agent is based on ZhiPu's self-developed models, including its latest reasoning models GLM-Z1-Air and GLM-4-Air0414. The company said that GLM-Z1-Air is a model that "matches the performance of DeepSeek's R1, while running up to eight times faster and requiring only one-thirtieth of the computing resources of its counterpart".
AutoGLM will become open source on April 14, a move that aims to promote the development of the AI industry ecosystem, according to the company.
The launch follows a surge in Chinese AI products this year. In early March, Beijing-based start-up Monica unveiled Manus, a general AI agent that it claims can make decisions and execute tasks autonomously, with much less prompting required compared with AI chatbots like ChatGPT and DeepSeek. [Read my post "Manus, Another DeepSeek Moment" published March 10, 2025].
Currently, one hundred and sixty-two aircraft of this model have been delivered, with a total of 645 routes in operation, connecting 158 cities, and safely transporting more than 20 million passengers.




No comments:
Post a Comment