Anthropic’s New Claude Opus 4.5 AI Model Is Designed for Coding and Office Work

Anthropic’s newest version of its most powerful generative AI model could upend how you manage your spreadsheets. The company said Claude Opus 4.5, announced Monday, is aimed at things you do on the job, like coding and office work.

Google unveiled its powerful new Gemini 3 model last week, and OpenAI released GPT-5.1 the week before. Now it’s Anthropic’s turn. The company, which is popular with businesses and software workers, said Opus 4.5 is focused on getting work done, not generating content.

Claude Opus 4.5 will be available everywhere and will be a default model for Pro (starting at $17/month), Max (starting at $100/month) and Enterprise users.

Opus 4.5 is built to produce documents, spreadsheets and presentations and can automate menial office tasks by using your computer and browser. That includes its deployment in Claude for Chrome, a browser extension that lets Claude do internet tasks for Max users.

This release puts all three Claude models in the 4.5 generation. Anthropic released Sonnet 4.5, its midlevel model, in September and Haiku 4.5, its smallest model, in October.

Watch this: Google Debuts Gemini 3, AI Agents Make Poor Freelancers, Amazon Smart Delivery Glasses | Tech Today

02:51

Advanced reasoning models like Opus are designed to handle complex, demanding tasks. While a smaller, cheaper large language model will provide an answer based on the probabilities in its training data, a reasoning model will rerun and refine its operations to get a better or more complete answer. This takes longer, but it means the AI can handle more difficult operations.

Reasoning models are particularly useful for complicated programming projects or intensive research. The downside is they are slower and more expensive to run, which is why companies often restrict them to paid plans or have strict limits on usage.