The latest developments, announcements, and breakthroughs from Anthropic's Claude AI assistant.
Claude 5's Extended Thinking mode breaks through what researchers thought was a 2-3 year milestone, exceeding the previous record by 8.1 percentage points in genuine scientific reasoning.
Read full article →Anthropic shifts focus from static inference to dynamic orchestration with four adjustable thinking effort levels and expanded context window.
Read more →New integration enables continuous conversations across open files with preloaded skills for financial analysis workflows.
Read more →New pricing structure makes Claude's longest prompts more accessible for enterprise use cases and long-running agents.
Read more →Claude Opus 4.6's new technology achieves 76% accuracy on MRCR v2 benchmark, dramatically improving long-context performance.
Read more →Get the latest Claude news delivered to your inbox. No spam, just important updates.