Claude vs GPT Prompts
Model-specific patterns.
Prompt patterns, jailbreaks, red-team techniques.
Prompt Engineering is one of the most rapidly evolving subdomains in AI. Tools we recommended six months ago have been replaced or open-sourced. We rerun benchmarks quarterly and publish the deltas — not just static rankings.
What you'll find here:
Every piece in this category has been reviewed by an editor with hands-on experience in the relevant tool or workflow. We don't publish AI-only content. We don't accept gifted hardware. We don't publish reviews of products we haven't used.
If we recommend a paid tool, the affiliate link is clearly marked with data-aff, and the recommendation is independent of payout. We've actively rejected several lucrative partnerships because the products didn't survive our internal testing.
Our test methodology is published per-review at the bottom of each comparison piece. In general:
Browse the 7 pieces in this category below. If you have a topic request or a correction, our about page has contact details.
Model-specific patterns.
From production.
What survives reasoning models.
The decision tree.
20 patterns that ship.
Json mode, tools, function calls.
The patterns that converge.