Anthropic has officially introduced Claude Opus 4.6, the most powerful AI model the company has released to date. The launch comes only months after Opus 4.5, highlighting the rapid evolution of Anthropic’s AI roadmap.
The most notable new feature is Agent Teams, which allows users to divide complex tasks among multiple AI agents working in parallel. Instead of a single agent handling tasks sequentially, each agent can take ownership of a specific part of a project and coordinate directly with others.
According to Anthropic, this approach dramatically increases productivity and mirrors the way high-performing human teams collaborate on large-scale projects.
Scott White, Anthropic’s Head of Product, described the feature as similar to having a team of skilled professionals assisting users simultaneously. Agent Teams are currently available as a research preview for API users and paid subscribers.
Claude Opus 4.6 also introduces a significantly expanded context window of up to one million tokens, matching the capabilities of Anthropic’s Sonnet models. This enables seamless processing of extremely long documents, large codebases, and complex datasets.
Another major improvement is native integration with Microsoft PowerPoint. Users can now create and edit presentations directly within PowerPoint using Claude, eliminating the need for exporting and importing files.
In performance benchmarks, Opus 4.6 delivered strong results. It scored 65.4% on Terminal-Bench 2.0 and 84.0% in Agentic Search, outperforming competing models such as Gemini 3 Pro and GPT-5.2.
The model also demonstrated significant gains in advanced reasoning benchmarks like Humanity’s Last Exam and ARC AGI 2, particularly in financial analysis, administrative automation, and agent-based workflows.
Anthropic notes that Claude Opus has evolved from a developer-centric tool into a broadly useful AI system now adopted by product managers, financial analysts, and professionals across many sectors.
