For years, AI tools have mostly helped people write emails, generate text, or answer questions. But what if an AI could open apps, work on spreadsheets, edit documents, and move your cursor just like a human user? That’s the idea behind GPT-5.4, a new AI model introduced by OpenAI.
The company says this model is built to handle complex professional work and can even operate a computer on the user’s behalf across different applications. The launch shows how AI is slowly moving from being just a conversation tool to becoming a digital assistant that can complete tasks inside software. AI that can actually use your computer One of the biggest highlights of GPT-5.4 is its native computer-use capability. According to OpenAI, the model can interact with a computer much like a person would. The system can send keyboard commands, move the mouse, and take screenshots to understand what is happening on the screen. By analysing what it sees, the AI can decide the next step and carry out actions in different programs. For example, it could potentially: This means AI could handle several routine tasks that people usually do manually on their computers. A step toward AI’s ‘digital agents’ The release of GPT-5.4 comes at a time when many technology companies are trying to build AI agents, systems that can complete tasks online or inside apps with minimal human input. Over the past year, several tools have appeared that can take control of a computer to perform tasks such as searching for products online or even ordering items automatically. GPT-5.4 is another step in this direction, bringing AI closer to becoming a fully functional digital assistant. Also read: Did you drop your phone in water this Holi? Try these simple hacks immediately to avoid any damage to your device
Introducing GPT-5.4 Thinking inside ChatGPT Alongside the main model, OpenAI has also launched a reasoning-focused version called GPT-5.4, “Thinking inside ChatGPT.” This version is designed to solve complex questions that require multiple steps of reasoning before reaching an answer. Inside ChatGPT, the model can even show an outline of how it is approaching a problem, helping users understand the steps the AI is taking. Another useful feature is that users can change their request while the response is being generated, instead of restarting the entire process. OpenAI explains: This makes it easier to guide the model toward the exact outcome you want without starting over or requiring multiple additional turns. Better integration with tools and APIs Another improvement is how the model interacts with external tools and web browsers. OpenAI says GPT-5.4 can call tools and APIs more accurately and efficiently, making it easier for the AI to perform tasks that require multiple software systems to work together. This capability is particularly useful for developers building AI-powered workflows or automation tools. Also read: Just got a new iPhone? Do this first: Activate these 5 features immediately for better security and improved battery
Where GPT-5.4 will be available OpenAI is rolling out GPT-5.4 across several of its products. Developers will be able to use the model through the company’s API and its AI coding tool Codex. Meanwhile, the GPT-5.4 Thinking model is being introduced in ChatGPT for users subscribed to Plus, Team, and Pro plans. The company is also launching GPT-5.4 Pro, which is designed for maximum performance when handling complex tasks. This version will be available through the API as well as for ChatGPT Enterprise and Edu users.
For now, GPT-5.4 Thinking is available in the ChatGPT web app and Android version, while support for iOS is expected to arrive soon.
The post appeared first on .
