Claude Opus 4.7

4 hours ago

Claude Opus 4.7 is announced as a hybrid reasoning model with improved performance in coding, vision, and complex multi-step tasks.
It features a 1M context window and is described as more thorough and consistent on difficult work, especially in professional knowledge work.
Opus 4.7 is available for Pro, Max, Team, and Enterprise users on Claude, and for developers via Claude Platform, Amazon Bedrock, Google Cloud’s Vertex AI, and Microsoft Foundry.
Pricing starts at $5 per million input tokens and $25 per million output tokens, with cost-saving options like prompt caching and batch processing.
Key use cases include advanced coding for production-ready code, AI agents for complex multi-tool workflows, and enterprise workflows for multi-day projects.
The model uses adaptive thinking to adjust effort based on task complexity, spending more time on harder problems.
Benchmarks and customer testimonials highlight significant improvements over Opus 4.6, including better coding resolution, reliability, and efficiency.
Notable gains are reported in areas like coding benchmarks, agentic workflows, document reasoning, and multimodal understanding.
Safety and reliability are emphasized through extensive testing and evaluation, meeting Anthropic's standards.
Frequently asked questions address model recommendations for demanding use cases and pricing details.

Hasty Briefsbeta