According to multiple leaks, ChatGPT could soon offer a new version of its built‑in image generator — one that lets users choose from “preset styles” instead of having to manually describe the style in the text prompt.
This potential update appears to borrow from the features introduced in OpenAI’s video and image‑creation app Sora, indicating a convergence of creative tools across OpenAI’s ecosystem.
A screenshot shared by an engineer on X suggests a new “ImageGenV2Banner” in ChatGPT web version, with a message reading: “Make images with ChatGPT – Feature your friends, faster image generation, and more consistent editing.”
Users on web and mobile versions have reportedly seen visual cards for trending art styles (e.g. “Crayon,” “Chibi,” “Nightcam,” “Blockart,” etc.), which users can select before generating an image.
If this feature is rolled out, users may no longer need to painstakingly describe art style in text prompts — the preset styles would automatically apply the visual tone.
Earlier in 2025, OpenAI replaced its older image‑generation model with the multimodal GPT‑4o, which already allows native image creation and editing within ChatGPT.
At the same time, OpenAI launched Sora — a separate tool for AI‑generated video and image content — signalling a broader push to integrate AI‑driven creative tools across formats.
With preset styles, users might get more consistent visual outputs — no need to manually specify style details in each prompt. This could especially help creators and casual users who struggle with style descriptions.
Selecting a style from a menu or card is faster than writing descriptive sentences about art style. This streamlines the creative process and lowers the barrier for generating high-quality images.
With multiple preset options (like Crayon, Chibi, Blockart), users can experiment with different art forms easily — leading to more varied and creative outputs without needing deep prompt mastery.
The feature is currently visible only to some users (on web and mobile), in what appears to be a limited or beta rollout.
The company — OpenAI — has not yet officially confirmed that the feature will be globally available or exactly when.
Details about how style presets will impact aspects like resolution, rendering time, or moderation policy are not yet disclosed.