OpenAI Unveils GPT-5.4 with Enhanced Reasoning, Coding, and Task Automation
Key Points
- OpenAI launches GPT-5.4, the latest large language model.
- New ability to interpret screenshots and control browsers via keyboard and mouse commands.
- Supports multi‑step workflows without human intervention.
- Improved factual accuracy and reduced false claims compared with earlier models.
- Introduces a “Thinking” mode that visualizes the model’s reasoning process.
- Longer context retention enhances coding assistance and complex task handling.
- Rolling out to ChatGPT web and Android users; iOS support forthcoming.
- Pro version available for enterprise and academic customers.
OpenAI announced the release of GPT-5.4, the latest version of its flagship AI model. The update brings notable improvements in reasoning, coding assistance, and real‑world task automation. New capabilities allow the model to interpret screenshots, control browsers, and issue keyboard and mouse commands, enabling multi‑step workflows that previously required human input. GPT-5.4 also offers stronger research abilities, longer context retention, and a “Thinking” mode that shows its reasoning process. The model is rolling out to ChatGPT users, the API, and enterprise customers, with a Pro version for high‑performance workloads.
Introduction
OpenAI introduced GPT-5.4 as the newest iteration of its large language model, highlighting advances in reasoning, coding, and task automation. The rollout spans ChatGPT, the API, and developer tools, with versions tailored for everyday users and enterprise workloads.
Direct Computer Interaction
One of the most significant changes is the model’s ability to interact directly with computers. GPT-5.4 can interpret screenshots, operate browsers, and issue keyboard and mouse commands, allowing it to complete tasks across multiple applications without human intervention. This capability supports complex, multi‑step workflows that previously disrupted user productivity.
Enhanced Research and Reasoning
The update improves the model’s capacity to conduct multi‑round information gathering, combining findings into clearer, structured answers. OpenAI describes GPT-5.4 as its most factual model to date, noting a reduction in false claims compared with its predecessor.
“Thinking” Mode
GPT-5.4 introduces a “Thinking” mode inside ChatGPT, designed for complex prompts. This mode displays a visible outline of the model’s reasoning, allowing users to adjust instructions mid‑response and guide outcomes without restarting the conversation.
Longer Context and Coding Support
The model retains information across extended workflows, making it especially useful for coding tools such as OpenAI Codex. Developers can rely on GPT-5.4 to automate large or time‑consuming development tasks.
Availability
GPT-5.4 is currently rolling out to ChatGPT users on the web and Android, with iOS support expected soon. OpenAI also offers a Pro version aimed at enterprise and academic customers that need maximum performance for complex workloads.