Figma: ChatGPT Images 2.0 Now Available in Figma

ChatGPT Images 2.0 Arrives Across the Figma Suite

Figma has integrated OpenAI's latest image generation model, ChatGPT Images 2.0 (powered by gpt-image-2), across its full product lineup, making it available starting April 21, 2026. The model is now accessible through the Make Image and Edit Image AI tools in Figma Design, Figma Draw, Figma Slides, Figma Buzz, FigJam, and Figma Weave.

What's New with ChatGPT Images 2.0

ChatGPT Images 2.0 represents a meaningful generational leap over its predecessor. OpenAI's new gpt-image-2 model was specifically designed to handle the kinds of visual complexity that previously tripped up AI image generators.

Smarter, More Complex Visuals

The model excels at generating "smarter" images — compositions that involve structured visual logic rather than purely aesthetic creativity. This includes infographics, data visualizations, UI screenshots, magazine-style layouts, and illustrations with integrated text. Where older models routinely mangled text, menus, or dense compositions, gpt-image-2 renders these with dramatically improved fidelity.

Multilingual Text Generation

One of the most practically useful upgrades is enhanced support for non-Latin scripts. ChatGPT Images 2.0 renders Japanese, Korean, Chinese, Hindi, and Bengali characters with far greater accuracy than previous generations — a significant improvement for design teams working on globally localized products.

Face Preservation Across Generations

The model can preserve faces consistently across multiple image generations. This has direct implications for designers building brand assets, character systems, or presentation decks that require visual consistency across multiple generated outputs — no more re-prompting to maintain the same character across a set of images.

2K Resolution and Flexible Aspect Ratios

The underlying gpt-image-2 model supports output up to 2K resolution and can generate images at flexible aspect ratios (from 3:1 landscape to 1:3 portrait), making it suited for a wide variety of design artifact types — from social media assets to presentations to wide-format backgrounds.

Where to Access It in Figma

The updated model slots directly into the existing AI image tooling Figma users already know. Designers can access gpt-image-2 through:

• Make Image — available in Figma Design, Draw, Slides, Buzz, and FigJam
• Edit Image — available in the same product surfaces
• Figma Weave — for building repeatable, scalable AI image workflows on an open canvas

There are no workflow changes required; the model upgrade is transparent to the user. Figma's help documentation covers the specific AI credit consumption for gpt-image-2 usage.