Claude Major Update: Opus 4.7 with Self-Verification & 1M Context Now Live

Anthropic has recently pushed a series of important updates for the Claude model family, covering the launch of the flagship Opus 4.7, the default activation of 1M-token context windows, and the opening of Claude Cowork to enterprise teams. These new capabilities take Claude to the next level in coding power, long-text handling, and team collaboration.

Opus 4.7 Arrives – Coding and Self-Verification Upgraded

Anthropic’s Claude Opus 4.7 is now the strongest flagship model, with significant improvements in software engineering. It introduces a self-verification feature that automatically double-checks its work before outputting results, sharply reducing error rates.

Opus 4.7 also enhances visual understanding, tripling image resolution for more accurate processing of complex charts and documents. Notably, Opus 4.7 is priced the same as Opus 4.6, with no price increase despite the performance boost. It is now available across Claude’s product line as well as AWS, Azure, GCP, and other platforms.

1M Context Window Enabled by Default – A New Standard for Long-Text Processing

The 1M-token context window of Opus 4.6 is now enabled by default for Max, Team, and Enterprise subscribers – previously it was an optional feature. This massive context window allows Claude to process document-length content comparable to an entire novel in one pass, making it especially useful for legal contract review, scientific paper analysis, and similar scenarios.

At the same time, Opus 4.6 defaults to a medium-investment mode for Max and Team users, balancing response speed and reasoning quality. For deeper reasoning, users can switch to high-investment mode using the ultrathink keyword. Anthropic has removed Opus 4 and 4.1 from the first-party API, guiding users toward the improved versions.

Claude Cowork Opens for Enterprise – AI Agents Move to Team Collaboration

Research preview of Claude Cowork is now available for the Team and Enterprise plans, giving enterprise teams access to a desktop AI agent. Anthropic has released three new Cowork demo videos, covering an overview as well as Managed Agents applications targeting the legal and financial industries.

Managed Agents lets enterprises configure specialized AI assistants for their industry needs – for example, legal teams can customize a contract review agent, while financial teams can set up a data analysis assistant. These demos have been embedded in the official Guide pages, making it easy for users to understand the deployment process.

From Opus 4.7’s self-verification capability, to the default 1M context window, to the expansion of Cowork for enterprise teams, Claude is gradually building out a product matrix that spans from individual creation to team collaboration. If you’re looking for an AI assistant that balances code quality with long-text processing, these new features are worth a try.