OpenAI has launched its new GPT-5.3 Instant model, which it claims provides more accurate, contextual, and less wordy responses than its predecessor. Key improvements include a significant reduction in hallucination rates, fewer unnecessary refusals or moralizing preambles, and more directly usable information from web searches.
The model is designed to be more conversational and useful by avoiding overly cautious dead ends that characterized the previous GPT-5.2 Instant. It is now available to all users, while GPT-5.2 Instant will be retired for paid users in June 2026.
Main topics: The launch and key features of GPT-5.3 Instant; specific improvements over GPT-5.2 Instant; reduction in hallucination rates; changes in response tone and refusal behavior; availability and rollout timeline.
ChatGPT parent OpenAI launched its latest model GPT-5.3 Instant on Tuesday. OpenAI says its predecessor GPT-5.2 Instant is the company's most-used large language model (LLM) so far.
âGPTâ5.3 Instant delivers more accurate answers, richer and better-contextualised results when searching the web, and reduces unnecessary dead ends, caveats, and overly declarative phrasing that can interrupt the flow of conversation,â the company wrote in a blog post.
Based on the feedback received from users, the latest model aims to make the conversation more contextual but less wordy, and also improve the tone and relevance.
In contrast to the more cautious GPT-5.2 Instant which refused seemingly insensitive prompts, GPT-5.3 Instant gives more direct answers by reducing unnecessary refusals and cutting âoverly defensive or moralising preamblesâ, resulting in fewer dead ends and more useful responses.
For context, GPT-5.2 Instant was launched in December 2025, at a time when sexually explicit content was on the rise and was drawing significant public attention, prompting renewed scrutiny of institutional oversight.
âMore broadly, GPTâ5.3 Instant is less likely to overindex on web results, which previously could lead to long lists of links or loosely connected information,â OpenAI added.
This means it will deliver âimmediately usableâ information instead of summarising online data without putting active reasoning into the answer.
GPTâ5.3 Instant also claims to reduce hallucination rates by 26.8% when using the web and 19.7% when relying only on its internal knowledge, compared to GPT-5.2. Hallucination rate is the percentage of time an LLM produces incorrect or fabricated information as fact. A lower rate means fewer hallucinations.
The updated model is available to all users including free and paid users. âUpdates to Thinking and Pro will follow soon. GPTâ5.2 Instant will remain available for three months for paid users in the model picker under the Legacy Models section, after which it will be retired on June 3, 2026,â the company added.
âGPTâ5.3 Instant delivers more accurate answers, richer and better-contextualised results when searching the web, and reduces unnecessary dead ends, caveats, and overly declarative phrasing that can interrupt the flow of conversation,â the company wrote in a blog post.
Based on the feedback received from users, the latest model aims to make the conversation more contextual but less wordy, and also improve the tone and relevance.
In contrast to the more cautious GPT-5.2 Instant which refused seemingly insensitive prompts, GPT-5.3 Instant gives more direct answers by reducing unnecessary refusals and cutting âoverly defensive or moralising preamblesâ, resulting in fewer dead ends and more useful responses.
For context, GPT-5.2 Instant was launched in December 2025, at a time when sexually explicit content was on the rise and was drawing significant public attention, prompting renewed scrutiny of institutional oversight.
âMore broadly, GPTâ5.3 Instant is less likely to overindex on web results, which previously could lead to long lists of links or loosely connected information,â OpenAI added.
This means it will deliver âimmediately usableâ information instead of summarising online data without putting active reasoning into the answer.
GPTâ5.3 Instant also claims to reduce hallucination rates by 26.8% when using the web and 19.7% when relying only on its internal knowledge, compared to GPT-5.2. Hallucination rate is the percentage of time an LLM produces incorrect or fabricated information as fact. A lower rate means fewer hallucinations.
The updated model is available to all users including free and paid users. âUpdates to Thinking and Pro will follow soon. GPTâ5.2 Instant will remain available for three months for paid users in the model picker under the Legacy Models section, after which it will be retired on June 3, 2026,â the company added.
Meridiano
