Claude Opus 4.5 arrives. Here’s what's new.

AI company Anthropic released its latest flagship AI model, Claude Opus 4.5, this week. Opus is considered one of the best AI models out there for developers looking to ramp up their coding output or to create AI agents.

So, what does the latest model, Opus 4.5, bring to the table?

As expected, Opus 4.5 performs exceptionally well on the various AI model benchmark tests. As TechCrunch points out, Opus 4.5 is the first model to score over 80 percent on SWE-Bench verified, which the outlet calls a "respected coding benchmark."

It seems like Opus 4.5 likely becomes the preferred AI model for vibe coders as well as experienced programmers with the way it performed on these tests. And Anthropic is further driving that by bringing additional upgrades to its Claude Code product. 

According to Anthropic, along with Opus 4.5, Claude Code will receive two more upgrades. Anthropic says that Claude Code's Plan Mode "now builds more precise plans and executes more thoroughly" and is now available in its desktop app, which allows users to run multiple sessions at once. Anthropic provides an example of having one AI agent fix a bug while another one researches GitHub.

In addition, Anthropic is more broadly rolling out two other products, Claude for Chrome and Claude for Excel, to more subscribers. Opus 4.5 will be able to showcase its multitasking prowess in these products as users can provide AI agents with parallel tasks in their browser or in their spreadsheets.

Opus 4.5 also brings some memory upgrades to the AI model which will now give users the ability to chat without limits. Clause will now summarize earlier conversations and will allow users to continue a single chat thread for as long as they'd like.