GPT-5.4 is OpenAI's frontier model designed to bring together recent advances in reasoning, coding, and agentic workflows into a single system. It inherits the industry-leading coding strengths of GPT-5.3-Codex while improving how it operates across tools, software environments, and everyday professional tasks involving spreadsheets, presentations, and documents. The result is a unified model aimed at getting complex real work done accurately and with fewer iterations, rather than a narrow specialist. Its release covers ChatGPT (as GPT-5.4 Thinking), the API, and Codex, with a GPT-5.4 Pro variant offered in ChatGPT and the API for users who want maximum performance on complex tasks.

Several practical capabilities distinguish GPT-5.4 from earlier releases. In ChatGPT, GPT-5.4 Thinking can present an upfront plan of its reasoning, so users can adjust course mid-response and arrive at a more aligned final output without extra turns, and it shows stronger deep web research for highly specific queries while retaining context over longer chains of thought. GPT-5.4 is also the first general-purpose OpenAI model with native computer-use capabilities in Codex and the API, allowing agents to operate computers and run complex workflows across applications. GPT-5.4 Pro is documented with a roughly 1.05 million-token context window, image input alongside text attachments, and support for reasoning, tool calling, and external function use, making it well suited for professional automation, multi-step analysis, and code-heavy agentic pipelines.