Introducing Opus 4.7

Opus 4.7 is here

We just added Claude Opus 4.7 to the model lineup in stagewise.

This is Anthropic's most capable model to date. It handles complex reasoning, large-scale architectural decisions, and long multi-file tasks with noticeably more depth than its predecessor. The jump from Opus 4.6 to 4.7 is significant - especially for work that benefits from extended thinking.

What changed

• 1M token context window - up from 200k on Opus 4.6. That is a 5× increase, which means fewer truncation issues on large codebases and long conversations.
• Adaptive extended thinking - the model decides how much reasoning effort to apply based on the task at hand. For straightforward edits it stays fast; for deeper problems it takes its time.
• Same pricing tier - $5 per million input tokens, $25 per million output tokens. No price increase over Opus 4.6.

When to use it

Opus 4.7 is the right pick when you need the model to hold a lot of context at once - large refactors, cross-file changes, or conversations that have built up significant history. For lighter, faster work, Sonnet 4.6 and Haiku 4.5 remain solid choices.

The previous Opus 4.6 is still available for anyone who prefers it.

Try it now

Switch to Opus 4.7 from the model selector in stagewise. No setup needed - just pick it and go.