OpenAI's GPT-5.5 reclaims top AI benchmark ranking but costs 20% more to use via API

THE DECODER · April 24, 2026

AI Summary

•OpenAI released GPT-5.5, a new version of its flagship AI model that scored highest on standard industry benchmarks (tests measuring accuracy, reasoning, and knowledge). The API pricing increased 20% compared to the previous version.
•Despite topping benchmarks, GPT-5.5 still produces false information frequently — a persistent problem across all AI models where the system generates confident-sounding but incorrect answers. Among paid AI services, it remains cost-competitive relative to competing models from other companies.
•For developers and companies using OpenAI's API to build chatbots, customer service tools, or content systems, the 20% price increase means higher operating costs, but the improved benchmark scores suggest more accurate and capable outputs — forcing a choice between the price hike and potential quality gains for their end users.

Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.