Skip to content

OpenAI · AI model guide

GPT Image 2.5 for Architectural Rendering

OpenAI's GPT Image 2.5, offered in RenderChat as two variants, Sunburst and Flare. Both run as native conversations, so every follow-up message keeps working on the same image thread. Sunburst is RenderChat's default model.

Maker
OpenAI
Variants
Sunburst · Flare
Output
1K
Aspect ratios
Auto + 10 fixed ratios
References
Supported
Follow-ups
Continues the conversation
Plan
Free and Pro

Workflows

What you can render with GPT Image 2.5

Each guide covers how to prepare the input, a step-by-step workflow, example briefs and what to check in the result.

Tips

Getting good results from GPT Image 2.5

  • Start with Sunburst. Switch to Flare in the model menu when you want a second read of the same brief.
  • Write follow-ups as edits ("keep everything, warm the light") rather than restating the whole brief.

FAQ

GPT Image 2.5 questions

Is GPT Image 2.5 available on the free plan?

Yes. GPT Image 2.5 is available on the Free plan as well as Pro.

What resolution does GPT Image 2.5 render at?

GPT Image 2.5 renders at 1K in RenderChat, with auto + 10 fixed ratios.

Can I keep refining a GPT Image 2.5 result in chat?

Yes. GPT Image 2.5 runs as a native conversation in RenderChat, so each follow-up continues from the previous turn. Short edit instructions work well.

How do I select GPT Image 2.5 in RenderChat?

Use the start button on this page — it opens a new chat with GPT Image 2.5 Sunburst selected — or choose it from the model menu in the chat composer.

Render with GPT Image 2.5 in RenderChat.

Upload a model view, sketch or render, describe the result you want and refine it in chat.

Start with GPT Image 2.5

Cookies help us improve RenderChat. Accept to allow analytics and advertising cookies (Google, Meta), or decline to keep them off. Privacy