Anthropic Claude Opus 4.7 Enterprise Release
Key Questions
What is the performance of Claude Opus 4.7 on SWE-bench Pro?
Claude Opus 4.7 achieves 64.3% on SWE-bench Pro. It includes multimodal vision capabilities for code fixes and produces strict outputs.
How does Claude Opus 4.7 compare to 4.6 in terms of token usage and costs?
Claude Opus 4.7 produces significantly fewer output tokens than Opus 4.6, leading to reduced costs for a fair total cost comparison. This efficiency contributes to rising enterprise revenue.
What new features does Claude Opus 4.7 offer for enterprise use?
It supports multimodal vision for code fixes and strict outputs, with emerging agent insights. Enterprise revenue is rising due to these advancements.
Claude Opus 4.7 hits 64.3% SWE-bench Pro with multimodal vision for code fixes and strict outputs; fewer tokens vs 4.6 reduce costs. Enterprise revenue rising, agent insights emerging.
Sources (2)
Updated Apr 20, 2026