Anthropic has officially released Claude Opus 4.5, completing its 4.5 model family and positioning the Amazon and Alphabet-backed startup as a formidable leader in the AI race. The new flagship model is designed to compete directly with recent heavyweights from Google and OpenAI, boasting enhanced capabilities in coding, autonomous agents, and office productivity.
The headline achievement for Opus 4.5 is its prowess in software engineering. On the SWE-bench Verified benchmark, which simulates real-world coding tasks, the model scored 80.9%. This performance surpasses Google’s Gemini 3 Pro (76.2%) and OpenAI’s GPT-5.1 Codex Max (77.9%), solidifying its status as a top-tier tool for developers. Alex Albert, Anthropic’s head of developer relations, called it “the most intelligent model in the world for the things that we really care about.”
Anthropic is introducing features specifically aimed at business users. “Infinite Chats” allows the AI to maintain context across extended conversations, solving the “amnesia” problem common in long projects. Additionally, a new “effort parameter” lets developers tune the model’s performance; at “medium effort,” Opus 4.5 matches the performance of the faster Claude Sonnet 4.5 but uses 76% fewer tokens.
The release also includes practical integrations with Microsoft Excel and Google Chrome, enabling the AI to analyze spreadsheets and automate browser-based tasks. Early customers have reported significant efficiency gains, including a 20% improvement in financial modeling accuracy.
The launch comes as Anthropic gains significant ground in the business sector. A mid-2025 report from Menlo Ventures indicates that Anthropic has captured 32% of the enterprise AI market, overtaking OpenAI, whose share has reportedly dipped to 25%.
This growth is proving lucrative for its big-tech backers. Amazon and Google, having invested $8 billion and $2 billion respectively, recently reported massive unrealized gains from these stakes—Amazon recognizing a $9.5 billion pretax gain and Alphabet reporting $10.7 billion in net gains.
Claude Opus 4.5 is available immediately to Pro users starting at $17 per month and Max users at $100 per month. It also arrives as the default model for the coding agent in GitHub Copilot during a promotional period.
Your First 10 AI Skills
10 practical AI skills, copy-paste prompts and a 7-day plan to start using AI with confidence.
Download the guide →