IBL News | New York
Anthropic upgraded its smartest model, enabling it to sustain agentic tasks for longer, operate across larger databases, and feature better code review and debugging to catch its own mistakes.
This is Anthropic’s first Opus-class model, Opus 4.6, which features a 1-million-token context window in beta, meaning the system can recall more information per user session.
However, its most notable addition is the inclusion of “agent teams” in Claude Code. It means thatusers can now assemble agent teams to work on tasks together.
Claude Opus 4.6 is available on claude.ai, the API, and all major cloud platforms. For developers, pricing through the Claude API remains the same at $5/$25 per million tokens.
Opus 4.6 can apply its improved abilities to running financial analyses, conducting research, and creating and using documents, spreadsheets, and presentations.
Within Cowork, where Claude can multitask autonomously, Opus 4.6 can leverage all these skills.
“The model’s performance is state-of-the-art on several evaluations,” according to the
Anthropic made substantial upgrades to Claude in Excel and released a research preview of Claude integrated in PowerPoint as an accessible side panel.
It means the presentation can be crafted in PowerPoint with direct help from Claude.