Yazılar

Zhipu AI Launches Free AI Agent, Heats Up China’s Tech Race

Chinese AI startup Zhipu AI has unveiled a free-to-use AI agent named AutoGLM Rumination, further intensifying the fast-growing artificial intelligence competition within China’s tech industry. The announcement was made by CEO Zhang Peng during a launch event in Beijing on Monday.

AutoGLM Rumination is capable of executing complex tasks such as deep research, web browsing, travel planning, and writing research reports. It is powered by Zhipu’s proprietary models — the reasoning model GLM-Z1-Air and the foundation model GLM-4-Air-0414. According to the company, GLM-Z1-Air rivals DeepSeek’s R1 in output quality but operates up to eight times faster, while demanding significantly less computing power — just one-thirtieth of the resources.

AI agents like AutoGLM are designed to autonomously perform tasks and make decisions, and their popularity is rapidly rising as firms strive to commercialize AI tools in practical, real-world settings. The move by Zhipu comes on the heels of Manus launching what it claimed was the world’s first general AI agent — albeit at a premium price of up to $199 per month. In contrast, Zhipu is offering its agent completely free via its official website and mobile app.

Founded in 2019 as a spinoff from a Tsinghua University laboratory, Zhipu AI has rapidly gained momentum and recognition. Its GLM series of large language models, particularly GLM4, are reported by the company to outperform OpenAI’s GPT-4 on several benchmarks.

This latest product launch is buoyed by a wave of government-backed support, with the company securing three rounds of funding in one month. The most recent came from the city of Chengdu, which invested 300 million yuan ($41.5 million) into Zhipu.

As the AI ecosystem in China accelerates, Zhipu’s free access model could prove disruptive — democratizing access to advanced AI tools while pushing other domestic rivals and global players to adjust their pricing and strategies.

OpenAI Unveils New Developer Tools Amid Growing Competition from Chinese AI Startups

OpenAI launched a suite of new tools for developers on Tuesday, designed to help build advanced AI agents using a few key application programming interfaces (APIs). These tools come at a time of increasing competition from emerging Chinese AI startups, which are gaining ground with innovative solutions and low-cost alternatives to industry leaders in the United States.

The new Responses API is a core feature of the update, available to all developers at no additional cost. It replaces the previous Assistants API, which OpenAI plans to phase out by mid-2026. The Responses API offers enhanced capabilities for creating AI agents, which are designed to perform complex, real-world tasks independently of direct human intervention.

This release comes as Chinese AI startups like Monica have gained significant attention. Monica’s autonomous AI agent, Manus, which was launched recently, has been touted as being on par with or even superior to OpenAI’s DeepResearch agent in terms of performance. The company’s Manus AI has attracted praise for its ability to carry out tasks with minimal human involvement, which has generated significant buzz in the AI space.

Additionally, Monica has announced a strategic partnership with the team behind Alibaba’s Qwen AI models, marking another significant move in the growing competition between Chinese startups and established players like OpenAI. This partnership could further accelerate the development and capabilities of AI agents, pushing the industry forward.

As AI technology evolves rapidly, OpenAI’s release of the Responses API underscores its ongoing effort to stay ahead in the competitive landscape, offering developers more powerful tools to create innovative AI solutions. However, as Chinese startups continue to make strides, the battle for dominance in the AI space is heating up.