Anthropic has announced the arrival of Claude Opus 4.5, the latest iteration of its flagship AI model, rolling out state-of-the-art performance improvements in coding, computer automation, and office productivity tasks. The release marks a significant milestone in making AI practical for everyday business workflows, particularly through deep integration with Microsoft Excel and other productivity tools.
Excel Automation Reaches New Heights
The standout feature of Opus 4.5 is its dramatically improved ability to handle spreadsheet tasks. Early testers reported a 20 percent accuracy improvement on internal evaluations and a 15 percent boost in efficiency gains when working with Excel files. The model can now read complex spreadsheets, fix formulas, build financial models, and generate summaries with unprecedented reliability.
Claude for Excel, available to all Max, Team, and Enterprise users, brings the chatbot directly into Microsoft’s spreadsheet application as a sidebar tool. Users can upload files, clean messy data, create formulas, and perform complex analyses without leaving Excel. Support for pivot tables, charts, and file uploads comes built-in, making it practical for finance teams, analysts, and anyone drowning in spreadsheet work.
Expanding Beyond Excel
Alongside Opus 4.5, Anthropic is making several tools more widely available. The Claude for Chrome browser extension, previously in limited release, is now available to all Max users. This allows seamless integration of Claude’s capabilities directly into web workflows.
Perhaps the most user-requested feature arriving with Opus 4.5 is “infinite chat”—a capability that prevents the model from hitting context window limits. For paid users, conversations can now continue indefinitely across files and chats while maintaining consistency, eliminating a long-standing frustration in extended work sessions.
Performance Benchmarks and Competitive Positioning
Opus 4.5 has set a new performance standard across multiple benchmarks. Notably, it’s the first model to score over 80 percent on SWE-Bench Verified, a respected coding benchmark. The model excels across tool use benchmarks, general problem-solving tasks, and computer vision challenges.
The release comes hot on the heels of competitive moves from Google and OpenAI. Google launched Gemini 3 Pro on November 18, while OpenAI released GPT 5.1 on November 12. Anthropic claims Opus 4.5 is its safest model yet, with improved ability to reject prompt injection attacks—a critical security consideration for enterprise deployments.
For Developers and Enterprises
Developers can access Opus 4.5 immediately through Anthropic’s API and applications. Pricing starts at $5 per million input tokens and $15 per million output tokens for the new model. The release also includes improvements in agentic workflows, where the model can refine its own processes and manage groups of sub-agents more effectively—a critical capability for building autonomous AI systems that handle complex, multi-step business operations.
Anthropic has also partnered with Microsoft to integrate Claude models into Microsoft 365 Copilot, giving enterprise users additional options alongside OpenAI’s GPT models for powering their AI-assisted workflows.
The bottom line: Opus 4.5 represents a meaningful shift from generative AI to what some observers call “administrative AI”—systems that don’t just suggest ideas but actually handle the repetitive, detail-oriented work that drains productivity. For professionals spending hours on spreadsheets, financial modeling, and routine business tasks, Opus 4.5 aims to be the intelligent assistant that finally justifies the AI hype.