GPT-5.5 Instant makes ChatGPT more accurate while nixing ‘gratuitous emojis’

Source: 9to5Mac

ChatGPT should feel “smarter and more accurate” starting today, according to OpenAI. The company is replacing the default model with an update called GPT‑5.5 Instant and cutting back on “gratuitous emojis” in responses.

GPT‑5.5 Instant is the new default ChatGPT model

GPT‑5.5 Instant arrives on May 5 (5/5), two months after the last version upgrade to GPT‑5.3 Instant on March 3.

While the previous Instant model promised to make ChatGPT less cringe, the new version focuses on accuracy and conciseness.

OpenAI specifically calls out a departure from excessive emoji use:

With this update, the model’s responses are tighter and more to‑the‑point without losing substance, while keeping the warmth and personality that makes ChatGPT enjoyable to use. It can deliver the same information, often with more utility than previous models, while reducing verbosity and over‑formatting that can make responses too long. It also asks fewer unnecessary follow‑up questions and avoids clutter, such as gratuitous emojis.

Critically, OpenAI claims the model enhancement also leads to higher accuracy with fewer hallucinations (made‑up facts):

• Instant is now more dependable, with significant improvements in factuality across the board and the largest gains in domains where accuracy matters most.
• In internal evaluations, GPT‑5.5 Instant produced 52.5 % fewer hallucinated claims than GPT‑5.3 Instant on high‑stakes prompts covering medicine, law, and finance.
• It also reduced inaccurate claims by 37.3 % on especially challenging conversations users had flagged for factual errors.

GPT‑5.5 Instant replaces GPT‑5.3 Instant for all users starting today. Paid users will continue to have access to GPT‑5.3 Instant for three months.

Last month, OpenAI enhanced ChatGPT and Codex with the release of GPT‑5.5 Thinking and Pro, two models it calls “a new class of intelligence for real work.”

You can learn more about the latest from OpenAI’s new default ChatGPT model here.