Edit Your Image with Instruct: Make Smart Changes Using Text

Instruct lets you edit an existing image using natural language. Just upload or select a reference image, describe what you want to change, and the system will generate a new version based on your instruction.

You can use it to change the color of an outfit, remove an object, adjust lighting, switch the season, or make many other types of edits.

How to use it

Option 1 — Upload an image from your device

1. Click Reference image in the prompt panel

2. Upload your image

3. Set Mode = Reference

4. Choose Method = Instruct

5. Write your instruction in the prompt field

6. Choose a style model for editing. We recommend using Flux.1 Kontext or GPT Image for best results  

7. Click Create Image

Option 2 — Use an image from your project

1. Hover over an image in your project

2. Click the Remix button that appears

3. The image will load into the prompt panel automatically

4. Mode will be set to Reference and Method to Instruct by default

5. Write your instruction in the prompt

6. You can choose the original style model of the image, or use a model like Flux.1 Kontext or GPT Image for better control  

7. Click Create Image

Option 3 — Use the viewer

1. Click on any image to open the viewer

2. Write your instruction in the prompt

3. Click Create Image

This will apply the Instruct using the same model that generated the image. Mode will be set to Reference and Method to Instruct automatically.

How it works

The uploaded or selected image becomes the reference for generation. Your text instruction is used to make targeted changes to that image.

Flux.1 Kontext and GPT Image are the most accurate models for Instruct. They follow instructions precisely and preserve areas that don't need changes.

Other models — including base models and style models — also support Instruct mode, but often redraw the entire image while trying to keep the overall composition. This may result in images that look quite different from the original.

You can apply Instruct with any model — regardless of which model was used to generate the original image. For example, you can take an image made with a style model and apply Instruct using Flux.1 Kontext.

Make sure the aspect ratio of your reference image matches the generation frame. Otherwise, the image may be cropped or resized.

Model compatibility

All models support Instruct mode

Best results come from Flux.1 Kontext and GPT Image. These models apply edits more locally and keep the rest of the image intact.

Other models may re-render the whole image based on your prompt