Post #6192

🚀OpenAI launches GPT-5.4 with stronger reasoning and faster performance OpenAI has introduced GPT-5.4, alongside GPT-5.4 Thinking and GPT-5.4 Pro, now rolling out across ChatGPT, the API, and Codex. What’s new: • One frontier model: GPT-5.4 combines improvements in reasoning, coding, and agentic workflows into a single system. • More efficient: the model uses fewer tokens while delivering faster responses and improved factual accuracy. • Better deep research: GPT-5.4 Thinking improves long-form reasoning and deep web research capabilities. • Stronger context retention: it can keep track of information better when reasoning for longer. • Mid-response steering: users can now interrupt the model while it’s responding and add instructions to redirect it. Availability: • GPT-5.4, GPT-5.4 Thinking, and GPT-5.4 Pro are rolling out gradually starting today. • Available in ChatGPT, the API, and Codex. • Steering is launching first on Android and web, with iOS coming soon. AI interfaces are shifting from simple prompts to interactive control loops where users can guide the model while it’s thinking. https://openai.com/index/introducing-gpt-5-4/ @aipost🏴