The relentless velocity of the artificial intelligence arms race accelerated once again on Monday as Anthropic announced the immediate release of Claude Opus 4.5, a flagship model that the startup claims has eclipsed both human experts and rival systems from Google and OpenAI in complex software engineering tasks.
The launch marks the San Francisco-based lab’s third major model release in just two months, underscoring a frenetic pace of development that has propelled the company’s valuation to a staggering $350 billion.
The debut of Opus 4.5 serves as a direct counterstrike to recent moves by Anthropic’s primary competitors. According to the company, the new model is now state-of-the-art for “agentic coding”—the ability of an AI to autonomously plan and execute programming tasks. On SWE-bench Verified, the industry standard test set for measuring software coding abilities, Opus 4.5 reportedly outperformed Google’s Gemini 3 Pro, which was announced just last week, as well as OpenAI’s GPT-5.1.
Register for Tekedia Mini-MBA edition 19 (Feb 9 – May 2, 2026): big discounts for early bird.
Tekedia AI in Business Masterclass opens registrations.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Register for Tekedia AI Lab: From Technical Design to Deployment (next edition begins Jan 24 2026).
Perhaps the most striking metric of the model’s capability comes from Anthropic’s internal hiring data. The company revealed that when it tested Claude Opus 4.5 on the rigorous take-home exam given to prospective performance engineers, the AI scored higher than any human candidate in the company’s history. This level of proficiency marks the model as a highly specialized tool for professional software developers, financial analysts, consultants, and accountants—knowledge workers who require high-level reasoning rather than simple chatbot interactions.
Scott White, Anthropic’s product leader for Claude.ai, described the release as a milestone for users eager to expand their professional purview.
“The amount that we’re releasing to the market and the feedback loops that we’re generating from it just make me so unbelievably excited,” White said.
He noted that the ideal user for the new system is someone looking to push their creativity and build new things, as the model is “meaningfully better” at everyday enterprise workflows, such as conducting deep research and manipulating complex data in spreadsheets and slide decks.
The release creates a crowded timeline for the company, known for its “Claude” family of models. The startup unveiled its mid-sized Claude Sonnet 4.5 in late September, followed swiftly by the compact Claude Haiku 4.5 in October. The previous iteration of its flagship, Claude Opus 4.1, was released only in August. This rapid obsolescence cycle is fueled by massive capital injections; just last week, Microsoft and Nvidia announced multi-billion-dollar investments in the firm, cementing Anthropic’s status as a central pillar of the generative AI economy alongside OpenAI and Google.
Beyond the raw intelligence of the model, Monday’s announcement included a suite of integration updates designed to weave Claude deeper into corporate workflows. The company is expanding “Claude for Chrome,” a browser extension that allows the AI to take action across tabs, to all Max users. Additionally, “Claude for Excel,” which enables the AI to understand, edit, and analyze spreadsheets, is now generally available to all Max, Team, and Enterprise users. The startup is also bringing its specialized “Claude Code” environment to its desktop application, further lowering the barrier for non-engineers to generate software.
Claude Opus 4.5 is available immediately and will serve as the default intelligence engine for Anthropic’s Pro, Max, and Enterprise subscription tiers, signaling the company’s intent to capture the high-end enterprise market while the window of technological superiority remains open.