OpenAI Launching GPT-5.4The company says this is the latest version of its AI model that combines advances in reasoning, coding, and professional tasks involving spreadsheets, documents and presentations. It is also the first model of OpenAI with native computer usage capabilities, meaning it can operate the computer on your behalf and complete tasks in a variety of applications.
While OpenAI is bringing GPT-5.4 to its API and its AI-powered coding tool, Codex, it is bringing its reasoning model, GPT-5.4 Thinking, to ChatGPT. OpenAI says that GPT-5.4 can write code to operate a computer, as well as issue keyboard and mouse commands in response to screenshots. GPT-5.4 shows improvements when using a web browser, as well as its ability to more accurately and efficiently call tools and APIs to help complete tasks.
The model is better able to field questions that require it to gather information from multiple sources, as OpenAI says the model can “search continuously over multiple rounds to identify the most relevant sources, especially for ‘needle-in-a-haystack’ questions, and synthesize them into a clear, well-reasoned answer.” OpenAI claims that GPT-5.4 is its “most factual model yet”, with individual claims 33 percent less likely to be false than GPT-5.2.
Inside ChatGPT, GPT-5.4 Thinking will provide frameworks for its work for more complex queries, as well as allow users to edit or change their request during its response. “This makes it easier to guide the model toward the exact result you want without restarting or requiring many additional turns,” says OpenAI. The feature is now available on the ChatGPT web app and Android, but OpenAI says it’s “coming soon” to the iOS app.
GPT-5.4 is now available on ChatGPT, Codex, and API, with GPT-5.4 coming to Thinking Model Plus, Team, and Pro users. Along with the API, a GPT-5.4 Pro model for “maximum performance on complex tasks” is also available for ChatGPT Enterprise and Edu users.
