According to @OpenAI, GPT-5 is far less likely to hallucinate than previous models, indicating higher response accuracy and trustworthiness (source: OpenAI on X, Aug 7, 2025). According to @OpenAI, GPT-5 more accurately recognizes when tasks cannot be completed and communicates its limits clearly (source: OpenAI on X, Aug 7, 2025). According to @OpenAI, the post did not provide release timing, pricing, API availability, or any references to cryptocurrencies or tokens (source: OpenAI on X, Aug 7, 2025).