OpenAI releases GPT-5.5 Instant, a new default model for ChatGPT

Editor's take

OpenAI has introduced GPT-5.5 Instant as the new default model powering ChatGPT, emphasizing reduced hallucinations in specialized domains like law, medicine, and finance, while preserving its predecessor's rapid response times.

This move signals OpenAI's ongoing effort to bridge the gap between raw language generation and reliable, domain-specific application. For users and businesses relying on ChatGPT for critical information, this improvement addresses a long-standing concern about factual accuracy, potentially expanding its utility in professional contexts beyond general inquiry. It also underscores the competitive pressure to deliver not just powerful, but also trustworthy LLM performance.

Future developments will reveal the extent of these hallucination reductions in real-world, high-stakes scenarios. Key questions remain about the model's performance on diverse, niche legal statutes or complex medical diagnoses, and whether this iteration will necessitate further fine-tuning for specific industries or if it represents a robust, general improvement.