Introducing GPT-5.4

2026-03-05 · Source: OpenAI News · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems, Software Development & Engineering · Depth: Advanced, extended

Summary

OpenAI has released GPT-5.4 and GPT-5.4 Pro, their most capable and efficient frontier models designed for professional work, available in ChatGPT, the API, and Codex as of March 5, 2026. GPT-5.4 integrates advanced reasoning, coding, and agentic workflows, incorporating GPT-5.3-Codex's coding capabilities and improving tool and software environment interaction. Key enhancements include an upfront thinking plan in ChatGPT's GPT-5.4 Thinking, improved deep web research, and better context maintenance. In the API and Codex, GPT-5.4 is the first general-purpose model with native computer-use capabilities, supporting up to 1M tokens of context and introducing "tool search" for efficient tool integration. The model demonstrates significant performance gains across benchmarks like GDPval (83.0% wins/ties), SWE-Bench Pro (57.7%), and OSWorld-Verified (75.0%), while also being more token efficient and 33% less likely to produce false claims than GPT-5.2.

Key takeaway

For AI Architects and NLP Engineers building advanced agentic systems, GPT-5.4 offers significant improvements in computer use, coding, and tool integration. You should explore its 1M token context window and "tool search" feature to develop more robust, efficient, and less error-prone agents capable of complex, multi-application workflows. Consider migrating from GPT-5.2 to capitalize on its enhanced performance and reduced token usage for professional tasks, especially those requiring deep web research or visual debugging.

Key insights

GPT-5.4 enhances professional AI capabilities through integrated reasoning, coding, and agentic workflows with improved efficiency and safety.

Principles

• Integrate diverse AI capabilities for holistic performance.
• Prioritize token efficiency for cost and speed optimization.
• Implement steerability for user-guided AI interaction.

Method

GPT-5.4 employs a "tool search" mechanism to efficiently manage large tool ecosystems by dynamically appending tool definitions, reducing token usage and improving response times for complex, multi-step agentic workflows.

In practice

• Utilize GPT-5.4 for complex spreadsheet and presentation tasks.
• Develop agents with GPT-5.4 for computer automation via Playwright.
• Employ "tool search" in API for large tool ecosystems.

Topics

• GPT-5.4
• Large Language Models
• Agentic AI
• Computer Vision
• Code Generation

Code references

• openai/skills

Best for: CTO, AI Architect, NLP Engineer, AI Engineer, Machine Learning Engineer, Software Engineer

Related on AIssential

• OpenAI launches GPT-5.4 Thinking and Pro combining coding, reasoning, and computer use in one model — GPT-5.4 unifies coding, reasoning, and computer operation, significantly boosting performance in professional knowledge work and desktop navigation.
• I Tried The New GPT 5.5 And I’m Never Going Back — GPT 5.5 excels at complex task execution and agentic workflows, moving beyond simple query responses.
• OpenAI introduces GPT-5.4 with more knowledge-work capability — GPT-5.4 enhances agentic knowledge work through improved reasoning, context, visual understanding, and token efficiency.
• GPT-5.5 is HERE! — GPT 5.5 offers enhanced intelligence and token efficiency, particularly for coding and research, at a lower effective cost.
• Is GPT-5.4 Worth It? — GPT 5.4 integrates native computer use and enhanced reasoning, primarily benefiting agentic applications.
• GPT 5.4 First Test Results — GPT 5.4 significantly advances AI capabilities in computer use, professional tasks, and coding, setting a new performance benchmark.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by OpenAI News.