AB
AB
AiBoss

All Categories

project

AutoGLM - The world's first mobile agent launched by Zhipu AI

AutoGLM, launched by Zhipu, is the world's first mobile agent, providing users with efficient operational proxy services through AI technology. Based on the domestically developed GLM-4.5 and GLM-4.5V models, AutoGLM possesses inference, coding, and multimodal capabilities. Aut...

What is AutoGLM?

AutoGLM, launched by Zhipu, is the world's first mobile agent that provides users with efficient operational proxy services through AI technology. Based on the domestically developed GLM-4.5 and GLM-4.5V models, AutoGLM possesses inference, coding, and multimodal capabilities. AutoGLM executes in the cloud, without consuming local device resources. Users can use voice commands to perform various operations, such as ordering takeout, booking flights, and searching for properties. AutoGLM supports cross-device operation, overcoming hardware limitations. AutoGLM's operational execution capabilities are encapsulated as an API, allowing developers to integrate it into various hardware devices, such as AI glasses, watches, and home appliances.

On March 31, 2025, Zhipu launched its first free AI Agent with in-depth research and operational capabilities: AutoGLM.

Main functions of AutoGLM

  • Cloud Operation Agent:
    • Life SceneUsers can use voice commands to operate dozens of frequently used apps such as Meituan, JD.com, Xiaohongshu, and Douyin to complete tasks such as ordering takeout, booking flights, and searching for properties.
    • Office SceneAutoGLM can perform full-process workflows across websites, operating web versions of websites such as Lark, NetEase Mail, Zhihu, Weibo, Douyin, and Toutiao, completing information retrieval, content writing, generating videos, PPTs, or podcasts, and directly publishing them to social media platforms.
  • Independent operation:
    • Dedicated Smart Phone/ComputerAutoGLM equips AI with dedicated cloud devices, enabling it to operate independently without consuming resources on the user's local device, allowing the user to use other apps simultaneously.
    • Asynchronous proxy officeUsers can have AutoGLM continue to perform tasks even when offline, providing 24/7 uninterrupted service.
  • Cross-device operation:
    • Overcoming hardware limitationsAutoGLM can run on any device and in any scenario, supporting cross-device operation, such as mobile phones, computers, watches, glasses, home appliances, etc.
    • API encapsulationAutoGLM's operational capabilities are encapsulated as APIs, allowing developers to integrate them into various hardware devices, giving the hardware complete mobile-level operational capabilities.

AutoGLM's technical principles

  • Domestic model-drivenAutoGLM is based on Zhipu's latest open-source state-of-the-art language model GLM-4.5 and visual reasoning model GLM-4.5V, possessing reasoning, coding, and multimodal capabilities. The model can perform various tasks such as reasoning, coding, research, agentic and GUI operations, flexibly calling the most suitable "brain" to complete the execution according to the needs.
  • End-to-end asynchronous reinforcement learning:
    • ComputerRLThe paper proposes an API-GUI collaborative paradigm to improve data diversity and computational efficiency; it also improves GRPO and proposes an Entropulse mechanism to enhance exploration and strategy diversity.
    • MobileRLAn innovative difficulty-adaptive reinforcement learning method (inference bootstrapping warm-up + difficulty-adaptive GRPO) significantly improves the stability and convergence efficiency of mobile tasks.
    • AgentRLBy using cross-sampling and task advantage normalization mechanisms, the instability and uneven gradient distribution in multi-task training are resolved, thereby enhancing overall robustness and efficiency.

How to use AutoGLM

  • Online experience:
    • Visit the official websiteVisit the AutoGLM website: https://autoglm.zhipuai.cn/.
    • Start cloud machineOnce the page has loaded, click the Start button to launch the AutoGLM cloud device.
    • Interacting with AutoGLMEnter your question or instruction in the input box.
      • For example, enter the following command:
      • "Could you check my flight tickets from Beijing to Shanghai for tomorrow?"
      • "Order me a coffee."
      • "Help me generate a PowerPoint presentation about environmental protection."
    • View resultsAutoGLM processes requests in the cloud and returns the results to the page.
  • Client:
    • Downloadable applicationDownload the AutoGLM client.
    • Registration and LoginComplete account registration and login, then enter the application.
    • voice commandsAfter opening the application, click the voice input button, speak the command, and AutoGLM will automatically recognize the voice command and execute the corresponding operation in the cloud.
    • Text inputEnter the text command in the application's input box. After entering the command, click the "Send" button, and AutoGLM will perform the operation according to the command.
    • View resultsAutoGLM completes the task in the cloud and then returns the results to the device.

AutoGLM's project address

  • Official website address: https://autoglm.zhipuai.cn/
  • Project official website: https://xiao9905.github.io/AutoGLM/

Application scenarios of AutoGLM

  • Social media management: Automatically perform actions such as liking, commenting, and sharing on social media platforms.
  • Online shoppingSearching for products, comparing prices, placing orders, and tracking logistics on e-commerce platforms.
  • Travel bookingsSearch and book hotels, flights, train tickets, etc. on travel websites.
  • Takeout orderBrowse menus, place orders, pay, and track order status on food delivery platforms.
  • Daily Information InquiryFor example, you can check the weather, news, stock information, etc.
Back