OpenAI Rolls Out GPT-5.3 Instant to Boost Accuracy and Conversational Flow

OpenAI has launched GPT-5.3 Instant, an updated version of its most widely used AI model powering ChatGPT, designed to make everyday interactions more accurate, natural and helpful.

The new model emphasises real-world usability by reducing unnecessary caveats and overly cautious responses that users have flagged in earlier versions.

According to OpenAI, GPT-5.3 Instant delivers more reliably accurate responses and richer contextualised results when incorporating web information, while improving conversational flow and tone.

"This update focuses on the parts of the ChatGPT experience people feel every day: tone, relevance, and conversational flow. These are nuanced problems that don’t always show up in benchmarks, but shape whether ChatGPT feels helpful or frustrating. GPT‑5.3 Instant directly reflects user feedback in these areas," OpenAI said.

The update significantly reduces “hallucinations” by up to 26.8% with web access and 19.7% relying only on internal knowledge in high-stakes domains such as medicine, law and finance, based on internal evaluations.

The model also improves how it balances internet sources with its own reasoning, helping it surface key information upfront rather than defaulting to long lists of links or detached summaries. OpenAI said these changes make the model feel more fluid and conversational, addressing feedback about stiff or overly defensive phrasing in GPT-5.2 Instant.

GPT-5.3 Instant is now available to all ChatGPT users and developers via the API as gpt-5.3-chat-latest. Older Instant versions will remain accessible to paid users in a legacy section before retirement later this year. Future updates to other model variants, such as Thinking and Pro, are expected soon.

Analysts say the release reflects OpenAI’s focus on enhancing everyday AI interactions by prioritizing accuracy, tone and relevance—key factors that shape user experience beyond traditional performance metrics.