In this round of ChatGPT updates, the focus isn’t on “one more button,” but on rounding out the model lineup, workflows, and the mobile experience all at once. You’ll see the GPT‑5 series rolling out gradually, clearer model selection, and more “assistant-like” proactive updates. Below, I’ll break down the new ChatGPT features by use case and explain them clearly.
On the model side: gradual GPT‑5 rollout and improved response style
OpenAI has updated GPT‑5.2 Instant in ChatGPT and the API to improve response style and overall quality; related improvements will also gradually extend across the new-feature rollouts for ChatGPT plans. At the same time, GPT‑5 is being opened up gradually on the web, iOS, Android, and desktop, allowing more users to switch to the new model from the same entry point. For most people, the noticeable change from this update is that answers feel more consistent—more like something that’s both well-written and clearly explained.
Selector and settings: a more straightforward model entry for paid users
What many users care about most is “Which models can I actually choose?” After the update, 4o has returned to the model selector for paid users. Meanwhile, web settings now include a “Show additional models” toggle; turning it on reveals more available models (such as o3, o4-mini, 4.1, and GPT‑5 Thinking mini). Note that due to GPU constraints, 4.5 is available only to the Pro plan; this is also one of the clearer tier distinctions in these ChatGPT updates.
