GPT-5.5 Instant: OpenAI updates ChatGPT’s default model to reduce hallucinations

OpenAI has released GPT-5.5 Instant as the new default model for ChatGPT, replacing GPT-5.3 Instant for all users. The company says the update focuses on three areas: fewer factual errors, more concise responses, and better personalization.

According to OpenAI, internal evaluations show that GPT-5.5 Instant produced 52.5% fewer hallucinated claims than GPT-5.3 Instant on sensitive topics such as medicine, law, and finance. The model also reduced inaccurate claims by 37.3% on conversations that users had previously flagged for factual errors. Independent verification of these figures has not been published.

The new model also performs better on standardized benchmarks. TechCrunch reports it scored 81.2 on the AIME 2025 math test, compared to 65.4 for the previous version, and outperformed its predecessor on the MMMU-Pro multimodal reasoning benchmark.

On personalization, GPT-5.5 Instant can draw on previous conversations, uploaded files, and connected Gmail accounts to tailor responses. A new memory sources feature, available across all ChatGPT models, shows users which context informed a response and allows them to delete or correct outdated information. OpenAI states that memory sources remain hidden when a chat is shared with others.

For developers, the model is available via API under the name chat-latest. GPT-5.3 Instant remains available to paid users for three months before being retired.

Sources: OpenAI, The Verge, TechCrunch