GPT‑5.5 Instant - Smarter, more personal answers as ChatGPT's new default
by•
GPT-5.5 Instant replaces GPT-5.3 as ChatGPT's default model with smarter, more concise answers, improved personalization from past chats and connected Gmail, and memory source controls.
For ChatGPT users on all plans.
Replies
Best
Hunter
📌
GPT-5.5 Instant is now ChatGPT’s default model, replacing GPT-5.3 Instant globally.
It brings two major upgrades.
First, accuracy: hallucinated claims on sensitive medical, legal, and financial prompts dropped by 52.5% versus the previous default.
Second, a new transparency feature called memory sources. When ChatGPT personalizes responses using past chats, saved memories, files, or connected Gmail, it now shows exactly what context it referenced. Users can review, edit, or delete those sources.
Most AI personalization works like a black box. Memory sources changes that by adding a citation-like layer for personal context across all ChatGPT models, not just GPT-5.5 Instant.
Key features:
52.5% fewer hallucinated claims on high-stakes prompts
Memory sources with visible, editable context references
Personalization from chats, files, and Gmail
More concise responses by default
Better visual reasoning, STEM, math, and web search decisions
Available in API as chat-latest
Why it matters:
More reliable answers where accuracy matters most
Greater transparency and control over personalization
Less response bloat in everyday use
P.S. I hunt the latest and greatest launches in tech, SaaS and AI, follow to be notified →@rohanrecommends
Report
5.5 somehow feels like an actual improvement. It might be just me but I do feel like it's actually capable of more now.
One weird thing though, it seems like temporary chats now start mixing up with the main projects/chats and referencing them more. It's either just starting to do that, or it was way less obvious about it before.
Report
Making the smarter model the default is a bold move, most companies charge more for better. curious, if “more personal” means actual memory or just better context handling a session.
Report
Memory sources is the underrated feature here. Every personalized AI system has the problem of "why did it say that" — and the answer is usually buried in something the user can't inspect. Making personal context visible and editable is a harder UX problem than it looks. Curious how they handle conflicts when memories from different time periods contradict each other.
Replies
GPT-5.5 Instant is now ChatGPT’s default model, replacing GPT-5.3 Instant globally.
It brings two major upgrades.
First, accuracy: hallucinated claims on sensitive medical, legal, and financial prompts dropped by 52.5% versus the previous default.
Second, a new transparency feature called memory sources. When ChatGPT personalizes responses using past chats, saved memories, files, or connected Gmail, it now shows exactly what context it referenced. Users can review, edit, or delete those sources.
Most AI personalization works like a black box. Memory sources changes that by adding a citation-like layer for personal context across all ChatGPT models, not just GPT-5.5 Instant.
Key features:
52.5% fewer hallucinated claims on high-stakes prompts
Memory sources with visible, editable context references
Personalization from chats, files, and Gmail
More concise responses by default
Better visual reasoning, STEM, math, and web search decisions
Available in API as chat-latest
Why it matters:
More reliable answers where accuracy matters most
Greater transparency and control over personalization
Less response bloat in everyday use
P.S. I hunt the latest and greatest launches in tech, SaaS and AI, follow to be notified → @rohanrecommends
5.5 somehow feels like an actual improvement. It might be just me but I do feel like it's actually capable of more now.
One weird thing though, it seems like temporary chats now start mixing up with the main projects/chats and referencing them more. It's either just starting to do that, or it was way less obvious about it before.
Memory sources is the underrated feature here. Every personalized AI system has the problem of "why did it say that" — and the answer is usually buried in something the user can't inspect. Making personal context visible and editable is a harder UX problem than it looks. Curious how they handle conflicts when memories from different time periods contradict each other.
NoteThisDown
I find it a bit weird that there's every other day a openai launch about some new feature.
Can just follow OpenAI on their social media pages for this, don't understand how product hunt is the right channel for this.