Ruben Hassid · 7 min read · Gen AI Tools
Opus 5.5
Read the original on ruben.substack.com →
Claude's new Opus 5.5 model has reclaimed the top spot on AI benchmarks, requiring a fundamentally different prompting approach than previous versions. The article breaks down key changes in prompting best practices, highlighting that the new Claude thinks autonomously and responds better to specific, structured instructions rather than generic guidance prompts.
Key Takeaways
- Remove 'think harder' or 'think step by step' prompts — Opus 5.5 automatically determines its own reasoning depth without needing nudges
- Ban design clichés by naming exact patterns to avoid (e.g., cream backgrounds, pill-shaped buttons) rather than using vague terms like 'make it premium'
- Instruct Claude to 'explore broadly' before acting when it has access to connected tools like Gmail, Drive, or CRMs to prevent it from missing relevant context
- Use structured, output-specific prompts that define format, length, and deliverable rather than open-ended requests
- AI model releases are accelerating, with a new major model launching approximately every 18 days, making prompt strategy adaptation a continuous necessity
Excerpt
I know , you’re tired. ChatGPT beats Claude. Then Claude beats ChatGPT. And it’s accelerating. We now have a new model… every 18 days: As of today, September 27th, Claude is back at the top with its new “ Opus 5.5 ”. Claude is #1 on artificialanalysis.ai . Claude is the best AI. And Claude’s best model is Opus 5.5 (not Fable 5.1 - I know, it’s weird, but I believe they will soon launch Fable 5.5). The better you prompt it, the better results you get. Anthropic, Claude Parent’s company, just wrote a mega PDF on prompting 5.5. Prompting…