OpenAI Upgrades ChatGPT Default to GPT-5.5 Instant — 52.5% Fewer Hallucinations
Overview
OpenAI on May 5 replaced GPT-5.3 Instant with GPT-5.5 Instant as the default model across ChatGPT and the API (as chat-latest). The update delivers measurably sharper accuracy in high-stakes domains and cuts response verbosity by roughly 30%.
Key Improvements
- Fewer hallucinations: 52.5% reduction in hallucinated claims on medicine, law, and finance prompts
- Conciseness: 30.2% fewer words and 29.2% fewer lines — addresses over-explanation and excessive emojis
- Benchmarks: AIME 2025 math score 81.2 (up from 65.4); MMMU-Pro 76 (up from 69.2)
- Personalization: Enhanced context from past chats, files, and connected Gmail — rolling out to Plus and Pro users first
Transition
GPT-5.3 Instant remains accessible to paid subscribers for three months via model settings. The chat-latest API alias now points to GPT-5.5 Instant.
GPT-5.5-Cyber Preview
A limited preview of GPT-5.5-Cyber also launched, targeting vetted cyber defenders protecting critical infrastructure.
Source: OpenAI Blog
Related Articles
OpenAI launched GPT-5.5 Instant on May 5, replacing GPT-5.3 Instant as the default ChatGPT model. The update cuts hallucinated claims by 52.5% on high-stakes prompts and adds personalized search across past conversations, files, and Gmail.
OpenAI launched a personal finance preview for ChatGPT Pro users in the US on May 15, powered by Plaid to connect over 12,000 financial institutions. Users get a live dashboard of spending, investments, and subscriptions analyzed by GPT-5.5.
GPT-5.5 Instant is now ChatGPT's default model, replacing GPT-5.3 Instant. The update cuts hallucinated claims by 52.5% on high-stakes medical, legal, and financial prompts, and adds Gmail-based personalization and memory-source transparency.
Comments (0)
No comments yet. Be the first to comment!