GPT-5.4 With Computer Use Capabilities¹

OpenAI has unveiled GPT-5.4, its most advanced frontier model designed for professional applications, which includes native capabilities for computer usage and enhanced reasoning skills.

This model is accessible through ChatGPT, the API, and Codex, with a Pro version available for optimal performance on intricate tasks.

Key Insights:

- The first general-purpose model equipped with native computer-use capabilities for managing computers and complex workflows.

- Accommodates up to 1 million tokens of context, allowing agents to plan, execute, and verify tasks over extended periods.

- Achieves an 83.0% match or exceed rate against industry professionals on the GDPval benchmark, an increase from 70.9% for GPT-5.2.

- Incorporates an 'adjust course mid-response' feature in ChatGPT, enabling users to redirect the model's thinking while it is processing.

- The most token-efficient reasoning model to date, utilizing significantly fewer tokens than GPT-5.2, resulting in faster processing speeds and lower costs.

The Bigger Picture: The computer-use capabilities of GPT-5.4 signify a substantial advancement towards genuinely autonomous AI agents capable of independently navigating software environments. Coupled with its exceptional performance in professional knowledge work—matching human experts 83% of the time—this positions AI as a true productivity partner across various fields, rather than merely a text generator.