Anthropic Launches Opus 4.5, Setting New AI Benchmark with Unprecedented Coding Skills

Anthropic has released its most powerful AI model yet. The new Claude Opus 4.5 launched on Monday. It represents the final model in the company’s 4.5 series.

This release follows the earlier launches of Sonnet and Haiku. According to TechCrunch, the model demonstrates state-of-the-art performance across several key benchmarks. Its capabilities are a significant step forward for complex AI tasks.

Opus 4.5 Achieves Historic Coding Benchmark Score

The model’s performance on coding tasks is particularly notable. Opus 4.5 is the first AI to score over 80 percent on SWE-Bench verified. This is a respected benchmark for evaluating software engineering capabilities.

It also excels in tool use and general problem-solving. Anthropic highlighted its advanced computer use and spreadsheet functions. These improvements are central to the model’s new agentic capabilities.

New Features Target Professional and Enterprise Users

Alongside the model, Anthropic is broadening access to companion products. Claude for Chrome and Claude for Excel are now more widely available. The Chrome extension is for all Max users, while the Excel tool is for Team and Enterprise tiers.

A major upgrade involves memory management for long conversations. The new “endless chat” feature prevents interruptions when the context window is full. The model now compresses its memory automatically without alerting the user.

These memory improvements are crucial for complex tasks. They allow Opus to act as a lead agent coordinating other AI models. This is vital for exploring large codebases and documents effectively.

The AI landscape is highly competitive right now. Opus 4.5 enters a market recently updated by rivals. OpenAI’s GPT 5.1 and Google’s Gemini 3 were both released this month.

Anthropic’s Opus 4.5 establishes a new high-water mark for AI performance, particularly in coding. Its advanced memory and agentic functions signal a shift towards more autonomous and complex AI applications. The frontier of artificial intelligence continues to accelerate at a remarkable pace.

Info at your fingertips

What is Opus 4.5’s most significant achievement?

It is the first AI model to score above 80% on the SWE-Bench verified coding benchmark. This demonstrates a new level of proficiency in software engineering tasks. The achievement sets a new industry standard.

Who can access the new Claude for Excel tool?

The Excel-focused model is available to Max, Team, and Enterprise plan users. The Chrome extension is available to all Max users. These tools were previously in a limited pilot program.

How does the “endless chat” feature work?

It allows conversations to continue seamlessly when the context window limit is reached. The model compresses its memory in the background without user intervention. This prevents chat interruptions for paid users.

What is the main focus of the memory improvements?

The upgrades enhance the model’s ability to manage long-context operations. This is critical for agentic use-cases where Opus directs other AI models. It allows for better navigation of large documents and code.

Which companies are Anthropic’s main competitors?

OpenAI and Google are the primary rivals in the frontier model space. OpenAI released GPT 5.1 on November 12th. Google launched its Gemini 3 model on November 18th.