Prompting Claude Opus 5.5

Claude Opus 5.5 generates output tokens over 30% faster than Claude Opus 5 while typically completing tasks with fewer tokens, making it more efficient for both speed and cost. The model shows particular strengths in agentic coding tasks, code review, and knowledge work, matching or exceeding Claude Opus 5's performance at higher effort levels while requiring fewer steps. Existing Claude Opus 5 prompts should work without modification, though Anthropic provides specific prompting patterns to optimize performance for various use cases.