Context Editing & Mixing

Edit your images by prompting simple instructions.

Key Features

• Edit images using natural language.

• Maintain character consistency across different prompts, angles, and outfits.

• Mix visual ideas from two reference images into one coherent output.

• Perform local edits to change specific parts of an image.

• Explore variations in lighting, clothing, style, and setting while keeping your subject intact.

• Progressive edits with Auto-refresh to refine your creation step by step with minimal visual drift.

---

Overview

Context lets you guide the image generation process using a combination of text and up to two reference images. This gives you powerful control over the result, whether you're editing an existing image, blending two ideas, or developing a consistent character across different scenes.

Note: Context editing uses next-gen editing & mixing models, available exclusively on the Fast Supernet.

How to use Context

In general, you'll use use Context editing in the following way:

1. Create an image using any model or method.

2. Set it as the Context reference image. There are two main ways to do this on Sogni Studio:

   1. On the top bar, click the flower icon. This will automatically switch the model to Qwen Image Edit, and set your image as reference.

   2. Select your preferred "Editing & Mixing" models from the Model Explorer, then drag and drop the image into the Context module.

(a)

(b)

1. Enter a simple instruction describing what you want to change in the prompt field and click "Imagine." For example:

Prompt: "now the monster is riding a motorcycle"

Context keeps the subject, clothing, and style intact, while applying the changes described. It also supports character consistency, letting you show the same subject from different angles, wearing different clothing, or in new situations, while keeping their identity and visual traits intact. You can change the lighting, setting, or even the artistic style, and the character stays recognizable across all variations.

1. To make further edits, set the last generated image as a reference, enter new editing instructions, and click "Imagine." For example:

Prompt: "now the monster is wearing goggles, and has a big smile"

Tip: Enable Auto-Refresh (next to the reference image) to make progressive edits by automatically setting the last generated image as the new reference.

Simple instruction prompts work best. Avoid repeating the original image description. Unless you're changing the style, leave the style field empty.

---

Need to make precise changes? Just describe what to adjust. Context editing supports local edits, so you can say things like:

"Make the castle a small house"

"Add a sailboat in the lake"

"Change the sky to a stormy sunset"

"Change the lamp for a TV"

"She is now smiling"

"Remove the necklace"

"Convert the sketch into an amazing photo"

The model will update just the parts you mention, keeping the rest of the image untouched.

---

Prompt: "playing guitar at a rock concert" — Style: "Inkpunk"

---

Prompt: "change style" — Style Preset: "Art Period: Impressionism"

---

Need help? Ask in our Discord!

For any questions regarding the use of Sogni, tips, suggestions, bugs, sharing art, discussing AI and life, or simply hanging out... Join our Discord community! Join Sogni Discord ✨