OpenAI Launches GPT-5.4 With Advanced Reasoning, Coding & Long-Context Capabilities

OpenAI has unveiled GPT-5.4, the latest upgrade to its flagship artificial intelligence model, introducing improved reasoning, coding capabilities and long-context understanding aimed at professional and enterprise use cases.

This comes just days after the launch of GPT-5.3 Instant, an updated version of its most widely used AI model powering ChatGPT.

GPT 5.4 is being rolled out across ChatGPT, the company’s API and its developer coding environment OpenAI Codex. OpenAI described the release as its “most capable and efficient frontier model for professional work.”

GPT-5.4 introduces a new “GPT-5.4 Thinking” mode in ChatGPT that outlines a plan before generating responses, allowing users to adjust the direction of the answer while it is being produced. The update also improves the model’s ability to conduct deep web research and maintain context when handling complex or multi-step queries.

The new model combines advances in reasoning and agentic workflows with the coding capabilities previously seen in GPT-5.3-Codex. According to OpenAI, GPT-5.4 is designed to work more effectively with software tools, connectors and professional productivity tasks such as spreadsheets, documents and presentations.

OpenAI also claims GPT-5.4 is its most factual model to date. On internal evaluations using prompts where users had previously flagged factual errors, the model’s claims were 33% less likely to be false and full responses were 18% less likely to contain errors compared with GPT-5.2.

Benchmark tests show GPT-5.4 outperforming its predecessor on knowledge-work tasks spanning dozens of occupations, achieving state-of-the-art results in professional productivity benchmarks.

The model is currently available to ChatGPT Plus, Team and Pro users, while developers can access it through OpenAI’s API. A higher-performance GPT-5.4 Pro variant has also been introduced for complex workloads requiring maximum performance.