OpenAI Rolls Out GPT-5.5 Instant as ChatGPT Default

OpenAI announced Tuesday that it has begun replacing ChatGPT's default model, GPT-5.3 Instant, with the upgraded GPT-5.5 Instant, rolling out the change to all users immediately.

The new model demonstrates significant improvements in accuracy, according to OpenAI's internal testing. It produced 52.5 percent fewer hallucinated claims than its predecessor on high-stakes prompts in medicine, law, and finance. In conversations previously flagged by users for factual errors, inaccurate claims dropped by 37.3 percent. Benchmark results further support these gains: on the AIME 2025 math competition, accuracy rose from 65.4 percent to 81.2 percent; GPQA, testing PhD-level science reasoning, improved from 78.5 percent to 85.6 percent; CharXiv for scientific chart reasoning increased from 75.0 percent to 81.6 percent; and MMMUPro for expert multimodal questions climbed from 69.2 percent to 76.0 percent. HealthBench scores also advanced, with the professional clinical version jumping from 32.9 to 38.4 points.

GPT-5.5 Instant maintains the low-latency "Instant" tier designed for everyday speed while delivering more concise responses. OpenAI stated the model reduces verbosity, using roughly...