Ranked for this profession and this task on the Redrob Leaderboard, September 2026, which is updated monthly. No list shows more than two models from one maker.

Each price is a typical month of this one task at the maker's published API price, before volume or caching discounts, in the currency of the page's language. More effort means more thinking, so it costs more on the same model.

Effort uses each maker's own scale: Claude runs Low to Max, ChatGPT Instant to Pro, Gemini Low to High. Check every citation and clause before you rely on it.