OpenAI GPT-5.4 brings native computer use, 1M context, and spreadsheet hooks
OpenAI released GPT-5.4 with native computer-use agents, a 1M-token context window, and new Excel/Sheets integrations, alongside SDK changes developers need to adopt. OpenAI’s new GPT-5.4 family (“Thinking” and “Pro”) emphasizes longer-horizon workflows and efficiency, with claims of up to 47% fewer tokens on some tasks and a native Computer Use mode that can issue mouse/keyboard events and operate via libraries like Playwright, plus direct spreadsheet integrations for Microsoft Excel and Google Sheets [VentureBeat](https://venturebeat.com/technology/openai-launches-gpt-5-4-with-native-computer-use-mode-financial-plugins-for). In ChatGPT, 5.4 Thinking targets complex, multi-step reasoning and exposes an interruptible plan preview for Plus/Team/Pro, while Pro is reserved for higher tiers [TechRadar](https://www.techradar.com/ai-platforms-assistants/chatgpt/chatgpt-just-got-another-brain-boost-with-gpt-5-4-thinking-and-its-built-for-bigger-more-complex-tasks). For developers, the openai-python v2.25.0 SDK adds gpt-5.4, a new Computer tool, and Tool Search, while deprecating the prompt_cache_key param and touching message phase fields—warranting quick dependency audits and integration tests [GitHub release](https://github.com/openai/openai-python/releases/tag/v2.25.0). See OpenAI’s guidance to refine prompts and model behavior as you evaluate long-context and agentic workflows [Prompt guidance](https://developers.openai.com/api/docs/guides/prompt-guidance).