Anthropic introduces Claude Opus 4.5, its most capable model yet, that can beat Gemini 3 Pro in coding

After Google, Anthropic has unveiled its most advanced model yet, featuring advancements in software engineering and agentic AI. The Claude Opus 4.5 model is now available through Anthropic’s apps, API, and all major cloud platforms, with prices starting at $5 for input and $25 for output per million tokens.

Survey

Taking to the blog post, the company stated that the Opus 4.5 is capable of delivering best-in-class performance on real-world coding evaluations, including the SWE-bench Verified, where it outscored all the competing frontier systems, notably surpassing Google’s Gemini 3 Pro in software engineering tasks. It also added that the Opus 4.5 solved complex, multi-system bugs more reliably than its predecessors and competitors, often completing challenges that were previously out of reach for Sonnet 4.5.

Add As A Trusted Source For Google.

Add as a preferred source on Google

Anthropic also stated that the model got the highest score ever recorded on its internal engineering take-home exam within the two-hour limit, exceeding any human candidate evaluated so far. Additionally, the model also demonstrated unusually strong agentic reasoning, with testers highlighting its ability to find creative paths through multi-step tasks.

Anthropic also claimed that Opus 4.5 is its most robustly aligned system to date. It showed the strongest resistance among frontier models to sophisticated prompt-injection attacks.

The company also announced upgrades to the Claude Developer Platform, including a new “effort” control for adjusting depth of reasoning, improvements to Claude Code, expanded desktop support, and better long-context handling across the Claude app, Chrome extension, and Excel integration. The company has also increased the usage limits for Opus 4.5 for Max and Team Premium users.

“Alongside Opus, we’re releasing updates to the Claude Developer Platform, Claude Code, and our consumer apps. There are new tools for longer-running agents and new ways to use Claude in Excel, Chrome, and on desktop,” the company stated in its blog post.

Follow Us

Ashish Singh

Ashish Singh is the Chief Copy Editor at Digit. He's been wrangling tech jargon since 2020 (Times Internet, Jagran English '22). When not policing commas, he's likely fueling his gadget habit with coffee, strategising his next virtual race, or plotting a road trip to test the latest in-car tech. He speaks fluent Geek. View Full Profile