GPT-5.3 Instant Rolls Out with Fewer Refusals and Better Accuracy

OpenAI

Mar 3, 2026 · Updated Apr 25, 2026

OpenAI is rolling out GPT-5.3 Instant to all ChatGPT users, delivering more accurate answers, fewer unnecessary refusals, and reduced preachy disclaimers. The update cuts hallucination rates by up to 26.8% and improves web search contextualization.

GPT-5.3 Instant, OpenAI's update to ChatGPT's most-used model, is rolling out to all users. The release focuses on tone, relevance, and conversational flow — improvements that don't always show in benchmarks but shape everyday usability. The model significantly reduces unnecessary refusals and tones down overly defensive preambles. On accuracy, it cuts hallucination rates by 26.8% with web search and 19.7% on internal knowledge. When using web search, users get sharper contextualization, better understanding of question subtext, and more consistent response tone.

This release directly addresses one of ChatGPT's most-cited frustrations: responses that hedge, moralize, or refuse answerable questions. A model that gives direct, confident answers without unnecessary caveats improves the everyday experience for all users.

GPT-5.3 Instant is available to all ChatGPT users now. Try prompts that previously triggered over-cautious responses to experience the improvements firsthand.

View the full update on openai.com

OpenAI

@OpenAIMar 3

GPT-5.3 Instant in ChatGPT is now rolling out to everyone. More accurate, less cringe. https://t.co/oJpXsp9TBc

1.2k

View on X

Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →

See all AI news & updates from OpenAI →

Keep reading

OpenAI Rolls Out GPT-5.5 Instant with Deep Gmail Personalization

OpenAI is rolling out GPT-5.5 Instant as the default model for all ChatGPT users, featuring a more concise and natural tone. The update integrates context from connected Gmail accounts and saved memories while adding a transparency layer to show which data sources influenced a response.

Arena.ai Ranks GPT-5.5 Instant as a Top Tier Conversational Model

ArenaMay 9

Arena.ai Ranks GPT-5.5 Instant as a Top Tier Conversational Model

Arena.ai added OpenAI's GPT-5.5 Instant to its blind evaluation leaderboards, revealing the model's performance across text, vision, and specialized professional categories. The results show the model excels in multi-turn dialogue but lags behind high-tier variants in raw reasoning and document analysis.

Lovable Reports GPT-5.5 Gains in Efficiency and Roadblock Resolution

LovableApr 24

Lovable Reports GPT-5.5 Gains in Efficiency and Roadblock Resolution

Lovable's early testing of GPT-5.5 shows the model requires 23.1% fewer tool calls while improving performance on complex technical builds. These results demonstrate a measurable leap in agentic reasoning, allowing AI to navigate difficult coding tasks with fewer errors at the same cost as previous models.

WarpApr 25

Warp Integrates GPT-5.5 to Boost Agent Speed and Efficiency

Warp added OpenAI's GPT-5.5 to its terminal-based AI agent, delivering a 40% speed increase over the previous model version. The update focuses on improving the responsiveness and cost-effectiveness of autonomous development loops like site migrations and code verification.