I Built GPT Image 2.5: AI Image Generation with Both Classic Prompts and a Chat Agent

Jesus Lopez — Thu, 10 Sep 2026 15:52:51 +0000

I recently built GPT Image 2.5, a web-based AI image generation and editing workspace:

Most AI image tools follow the same pattern: choose a model, write a prompt, select a size, click generate, and wait for the result. That workflow is still useful, so GPT Image 2.5 supports the classic image generation flow.

But I also wanted to explore a more natural way to create images: generating through conversation.

Two ways to create images

GPT Image 2.5 currently supports both workflows:

Traditional prompt-based image generation
Conversational image generation with an AI agent

The traditional mode is good when you already know what you want. You can write a prompt, upload a reference image, choose the model and output size, then generate or edit the image directly.

The AI agent mode is different. Instead of filling out everything manually, you can chat with the agent about what you want to create. For example, you can describe a product photo, a poster idea, a portrait style, or a marketing visual, and the agent helps turn that conversation into an image generation task.

This makes the workflow feel closer to working with a creative assistant:

You can start with a rough idea
Ask the agent to refine the direction
Adjust the style through conversation
Explore multiple visual concepts
Generate images without needing to write a perfect prompt from the beginning

Supported models and use cases

The workspace includes Flare, Sunburst, and GPT Image 2.

Flare is useful for fast creative exploration. Sunburst is better suited for more detail-focused editing. GPT Image 2 remains available for workflows built around the previous generation.

Some example use cases:

Product visuals
Portraits
Posters
Campaign concepts
Social media images
Reference-based image editing
Creative visual experiments

Why add an AI agent?

A lot of people do not naturally think in “prompt format.” They may know the result they want, but not how to describe it in a way that works well for an image model.

The agent helps bridge that gap.

Instead of forcing users to write a polished prompt immediately, the site lets them describe the idea conversationally. The agent can help shape the request, clarify the visual direction, and guide the generation process.

For developers and makers, I think this is an interesting direction: AI image tools do not have to be just prompt boxes. They can become interactive creative workspaces.

What I am improving next

The project is still early, and I am working on improving:

The chat-to-image workflow
Better prompt refinement
More image editing options
More examples and templates
A smoother model selection experience

I would love feedback from developers, designers, marketers, and creators who use AI images in real workflows.

Website: https://gpt-image-2-5.org/

I Built an AI Tool That Turns Flat Images into Editable Layers

Jesus Lopez — Mon, 31 Aug 2026 10:05:09 +0000

I Built an AI Tool That Turns Flat Images into Editable Layers

A flat image can look complete, but it is often difficult to edit.

If you want to move a product, remove a shadow, replace a background, edit text, or reuse part of a composition, you usually have to recreate the image structure manually in Photoshop.

That was the problem we wanted to solve with Image2Layer.

What Image2Layer does

Image2Layer uses AI to analyze a flat JPG, PNG, or WebP image and separate it into meaningful editable layers.

Depending on the image, the result may include:

Main subjects
Foreground objects
Product elements
Text and logos
Shadows
Decorative elements
Background areas
Reconstructed regions behind removed objects

Each layer keeps its transparency, so it can be moved, hidden, edited, or recomposed independently.

Export formats

After processing an image, you can export:

Transparent PNG layers
A ZIP package containing the layers
A layered PSD file for Photoshop and other creative workflows

The goal is not to recreate the original source file perfectly. Instead, Image2Layer provides a useful editable starting point from an image that would otherwise be difficult to work with.

Who is it useful for?

Some possible workflows include:

Retouching product photos
Creating new e-commerce compositions
Reusing poster and marketing assets
Preparing images for animation
Separating characters and objects from illustrations
Rebuilding backgrounds
Moving from an AI-generated image to an editable design

How it works

Upload a JPG, PNG, or WebP image.
Choose the layer detail level.
Add separation notes if certain elements are important.
Generate the layers.
Review and download the result as PNG, ZIP, or PSD.

You can try it here:

Try Image2Layer

What we are still improving

AI separation is not equally easy for every image. Fine details, transparent objects, reflections, complex typography, and heavily overlapping elements can still be challenging.

We are continuing to improve:

Layer boundaries
Background reconstruction
Text and logo separation
PSD export quality
Control over the desired layer structure

If you work with image editing, design, or creative automation, I would be interested to hear what kind of images you would want to split into layers.

DEV Community: Jesus Lopez

I Built GPT Image 2.5: AI Image Generation with Both Classic Prompts and a Chat Agent

Two ways to create images

Supported models and use cases

Why add an AI agent?

What I am improving next

I Built an AI Tool That Turns Flat Images into Editable Layers

I Built an AI Tool That Turns Flat Images into Editable Layers

What Image2Layer does

Export formats

Who is it useful for?

How it works

What we are still improving