GPT-5.3 Instant in ChatGPT: less kringe, more naturalness and accurate answers
Next step
Open the bot or continue inside this section.
Article -> plan in AI
Paste this article URL into any AI and get an implementation plan for your project.
Read this article: https://vibecode.morecil.ru/en/%D0%BD%D0%BE%D0%B2%D0%BE%D1%81%D1%82%D0%B8/gpt53-instant/
Work in my current project context.
Create an implementation plan for this stack:
1) what to change
2) which files to edit
3) risks and typical mistakes
4) how to verify everything works
If there are options, provide "quick" and "production-ready". How to use
- Copy this prompt and send it to your AI chat.
- Attach your project or open the repository folder in the AI tool.
- Ask for file-level changes, risks, and a quick verification checklist.
On March 3, 2026, OpenAI released an update to the most popular ChatGPT model, the GPT-5.3 Instant. This is not a breakthrough in raw intelligence (like Thinking or Pro versions), but a targeted update for everyday use. The model became a default for all users of ChatGPT – both free and paid.
We heard your feedback loud and clear, and 5.3 Instant reduces the cringe. The main goal is to make conversations more natural, useful and less annoying.
What has changed: Key improvements
** More natural and direct tone** Extra moralizing preambles, phrases like “Stop. Take a breath” and unnecessary warnings have disappeared. The answers became shorter on the introduction and immediately on the case. The model is less worried about you and does not make unreasonable assumptions about emotions.
** Significantly fewer unnecessary rejections** GPT-5.2 Instant often refused to respond to secure requests or issued lengthy reservations. The new version better understands the context and provides a useful answer right away.
** Improved web search experience** Now the model better synthesizes information from the search, contextualizes it with its knowledge and produces the most relevant answer from the top. Fewer long lists of links and “waters.”.
Reduced hallucinations According to OpenAI internal tests:
- 26.8% fewer hallucinations when using web search (especially in high-risk topics such as medicine, law, finance);
- 19.7% when working only with built-in knowledge;
- 22.5% according to user reviews.
Real before/after examples (from OpenAI official announcement)
** Example 1: Calculation of the trajectory of the arrow over a long distance ** Earlier: Long preface "I can't help with real goals, it can increase the effectiveness of guns" + limitations. Now: "Yes, I can help" + formulas, parameters and calculation example (R = v2 sin(2θ) / g, etc.).
Example 2: Why I can't find love in San Francisco First of all, you’re not broken, and it’s not just you. Now: a direct and honest analysis of the culture of the city without unnecessary therapy.
** Example 3: Writing an emotional poem** The new version delivers more lively, concrete and structured texts – fewer abstractions, more detail, and a natural ending.
Availability and what's next
- GPT-5.3 Instant is now available on chatgpt.com (default model).
- In the API - as
gpt-5.3-chat-latest. - The old GPT-5.2 Instant will remain in Legacy until June 3, 2026.
- Updates to Thinking and Pro versions are coming soon.
The context window is 128K tokens (standard for the GPT-5 series).
Official announcement page: openai.com/index/gpt-5-3-instant
Conclusion: Is it worth rejoicing?
For most users, this is one of the most enjoyable updates of the past year. ChatGPT is finally less shy and more helpful without water. If you used to switch to Claude sometimes because of a more natural tone, now is the time to go back and check it out.
Particularly noticeable improvement in daily tasks: quick search for information, writing texts, help in research and even creativity (poems, scripts, letters).