OpenAI Releases New GPT-5.4 Flagship Model

OpenAI is launching today GPT-5.4, its latest flagship model, which it considers as its “most capable and efficient frontier model for professional work.” The new model is available in two flavors in ChatGPT, the GPT‑5.4 Thinking reasoning model and the GPT‑5.4 Pro variant delivering maximum performance for complex tasks.

GPT-5.4 comes just two days after the release of GPT-5.3 Instant, which the company said was designed to improve “everyday usability.” As a reminder, OpenAI’s “Instant” models are optimized for low-latency and efficiency, while its “Reasoning” models are optimized for complex reasoning.

To be clear, GPT-5.3 Instant is now available for all ChatGPT users, while GPT‑5.4 Thinking is rolling out first to ChatGPT Plus, Team, and Pro users. It will be replacing GPT‑5.2 Thinking, which will still remain available for three months in the model picker. GPT‑5.4 Pro, however, is currently only available for ChatGPT users with Pro and Enterprise plans. The two GPT-5.4 models are also available in Codex, the new AI coding environment that’s now available as a native app on macOS and Windows.

“GPT‑5.4 brings together the best of our recent advances in reasoning, coding, and agentic workflows into a single frontier model,” OpenAI said today. “It incorporates the industry-leading coding capabilities of GPT‑5.3‑Codex⁠ while improving how the model works across tools, software environments, and professional tasks involving spreadsheets, presentations, and documents. The result is a model that gets complex real work done accurately, effectively, and efficiently—delivering what you asked for with less back and forth.”

In addition to advances in general reasoning, coding, and professional knowledge work, GPT-5.4 Thinking allows users to adjust its activity mid-response. The new reasoning model can also better maintain context for questions that require longer thinking, and it can more persistently search the web across multiple rounds.

OpenAI also pointed out that GPT-5.4 is its first general-purpose model with native computer-use capabilities. “It is excellent at writing code to operate computers via libraries like Playwright, as well as issuing mouse and keyboard commands in response to screenshots,” the company explained.

While OpenAI continues to release new AI models at a regular pace, the company was recently in hot water following its agreement with the US Department of War. While the ChatGPT maker claimed that its technology won’t be used for domestic surveillance, Anthropic’s Claude recently climbed to the top of the US App Store after the company was blacklisted by the Pentagon.