**GPT-5.5 Instant: The Updated Default ChatGPT Model**
OpenAI has declared that starting today, ChatGPT will appear “smarter and more precise” with the launch of the new default model, GPT-5.5 Instant. This enhancement, which is set to roll out on May 5, signifies a notable upgrade over the earlier version, GPT-5.3 Instant, introduced merely two months ago.
The key objective of GPT-5.5 Instant is to boost accuracy and brevity in replies. OpenAI has intentionally worked to limit the use of excessive emojis, striving for responses that are more concise and direct without compromising on content. The aim is to preserve the warmth and character that users appreciate while conveying information more efficiently.
OpenAI stresses that this model enhancement results in improved accuracy, with a significant decline in “hallucinations,” or errors in the information presented. Internal assessments have revealed that GPT-5.5 Instant generates 52.5% fewer hallucinated claims compared to its predecessor on high-stakes prompts, especially in crucial fields like medicine, law, and finance. Additionally, it has lowered incorrect claims by 37.3% in difficult discussions previously highlighted by users for factual inaccuracies.
The launch of GPT-5.5 Instant will supplant GPT-5.3 Instant for all users, though paying users will have continued access to the former version for an extra three months. This update follows recent advancements made to ChatGPT and Codex with the introduction of GPT-5.5 Thinking and Pro, which OpenAI refers to as “a new class of intelligence for real work.”
For further details on the newest updates from OpenAI and the features of the GPT-5.5 Instant model, you can check their official page.