GPT-5.1 represents a meaningful step forward in LLM capabilities. Three key improvements stand out:
1. Engine Segmentation & Personality Presets
The ability to segment different engine types with distinct personalities is genuinely useful. As a GTM builder, this means I can deploy contextually-optimized responses without extensive prompt engineering overhead.
2. Superior Instruction Following
The model now handles multi-step constraints simultaneously. Complex instructions that previously required 3-4 iterations now work on the first try. This directly reduces latency in production systems.
3. Improved Tone Adaptation
GPT-5.1 understands conversational context better. It shifts tone appropriately based on input, which matters more than people realize for enterprise adoption. Technical superiority loses to human-like interaction every time.
The Real Unlock: This isn't a revolutionary leap. It's a solid incremental advance that compounds when deployed at scale. The real advantage goes to teams building on top of this—not those claiming AGI is here.
Raycast
Definitely a popular use case — certainly seems like ChatGPT could become a superapp by year's end!
https://vimeo.com/1151655050/ba15f56a51?fl=pl&fe=cm
minimalist phone: creating folders
This can help with early diagnostics. I already used ChatGPT to describe the symptomps and it was pretty accurate in identifying my problem. My another favourite usecase is to use it for conversation in different languages I am trying to learn.
Vozo AI — Video localization
Is this a new GPT model, or is it just a conversation with specific instructions added?
Raycast
@josie_oy It's not a new model; it's a dedicated "space" and set of features (like ChatGPT Image Generation) powered by existing GPT models with added health-specific context, controls, and protections.
The “Health” space inside ChatGPT will layer on purpose‑built encryption/isolation, separate memories, and integrations (medical records, Apple Health, apps) to ground answers in your data.
Health chats aren’t used to train foundation models, and information in Health doesn’t flow back into non‑Health chats.
You can think of it as a specialized mode plus tooling and policy, designed with physician input and evaluated via HealthBench—supporting navigation and understanding, not diagnosis or treatment.
Does this plugin to wearables - I would love to connect my whoop to it
SigniFi
thats cool! hope oai make me live better
I have already seen people use ChatGPT for understanding symptoms and next steps. Having this framed clearly around health and wellness makes a lot of sense. Interested to see how this evolves.
Interesting to see how it evolves! I see that for now only users outside the European Economic Area, Switzerland, and the United Kingdom are eligible. I’d love to know whether the full release is expected to exclude those regions as well.