OpenAI Launches ChatGPT Images 2.0: New 'Thinking Ability' and Consistent Multi-Image Generation

OpenAI recently launched ChatGPT Images 2.0, an AI image generation tool built on the GPT Image 2 model. The core upgrade is the addition of 'thinking ability'—when enabled, the tool can retrieve information online, generate visual analysis content from uploaded files, and reason about image structure before generation. This feature is currently available to ChatGPT Plus, Pro, Business, and Enterprise subscribers.

The new version breaks through in multi-image generation: it supports up to 8 images per prompt and maintains consistency in characters, objects, and styles across different scenes. This will significantly lower the creation barrier for comic pages, social media series content, and interior design plans.

In terms of image quality, it supports up to 2K resolution, expands aspect ratios to 3:1 and 1:3, and optimizes styles like pixel art, comics, and movie stills. For multilingual support, it has improved text generation capabilities for Chinese, Japanese, Korean, Hindi, and Bengali.