OpenAI just released GPT-5.4!
From Jason West of ChatGPT Users: OpenAI just released GPT-5.4⦠and itās a pretty big step forward. If youāre using ChatGPT regularly, youāll probably start noticing a few differences straight away. Here are a few highlights: 1. Much stronger at real workGPT-5.4 is designed for professional tasks like writing documents, building spreadsheets, analysing information, and producing structured outputs. In testing, it performed as well as or better than industry professionals in many knowledge-work tasks. 2. Better coding abilityIt now includes the coding strengths from the latest Codex models. That means stronger debugging, better code generation, and more reliable help when building apps, scripts, or automations. 3. Can work with tools and software more effectivelyOne of the biggest improvements is how it works with external tools, apps, and workflows. This is particularly important for AI agents that need to interact with websites, spreadsheets, APIs, and other systems. 4. Longer memory and contextThe model can handle extremely large inputs (up to 1 million tokens in some environments), which means it can process large documents, projects, or codebases much more effectively. 5. Better at searching the webGPT-5.4 is significantly stronger at deep research tasks where it needs to find and combine information from multiple sources. 6. Fewer factual mistakesOpenAI says responses are about 33% less likely to contain incorrect claims compared with the previous generation. What this means for us For most people in this community, the biggest benefits will be: ⢠better prompts⢠more reliable outputs⢠stronger coding help ⢠improved research⢠fewer hallucinations If you build workflows, automations, prompts, or AI tools, this release should make things noticeably smoother. Have you tried GPT-5.4 yet? What differences have you noticed so far?