---
title: Edit an image without starting over
display: '*Edit* an image without starting over'
nav_title: Edit images
summary:
  - "Edit instead of regenerating when most of the image is already right: the rest of the image stays as it is."
  - "Describe the change in a sentence, paint a mask over the area, or extend the frame: pick the tool that fits the change."
  - "Every edit makes a new asset, so the original stays safe in your library."
description: Change one element by prompt, repaint a masked area, extend the frame or change the format, and check that nothing else moved.
hide_lede: true
section: images
order: 40
type: guide
level: Start
roles: [all]
time: { read: 9, do: 15, wait: 2 }
needs: ["An image in your library"]
last_verified: 2026-11-11
models_checked: ["Nano Banana 2.1", "Scenario Inpaint", "Photoroom Expand", "GPT Image 2.5 Sunburst", "Smart Reframe", "Reframe"]
hero: images/edit-images/robot-owl-prompt-edit.webp
related: [your-first-edit, remove-and-replace-backgrounds, upscale-and-enhance-images]
video: tutorials/edit-images.mp4
video_caption: Edit images
about: "When most of an image is right, edit it instead of starting over. This lesson takes one image, a moss-covered robot holding a fox cub in a misty river, and edits it three ways: Edit with Prompts with Nano Banana 2.1 to change one element or the whole scene, Retouch to repaint one small area under a mask, and Expand with Photoroom Expand to grow it into a wide banner. Each edit makes a new image, so the original stays safe. The models and prices in this lesson are as of November 2026."
chapters:
  - { at: 0, title: "Edit with Prompts", image: "/media/images/images/edit-images/ch-01-edit-with-prompts.webp" }
  - { at: 63.3, title: "Retouch", image: "/media/images/images/edit-images/ch-02-retouch.webp" }
  - { at: 88, title: "Expand", image: "/media/images/images/edit-images/ch-03-expand.webp" }
thumb: images/edit-images/thumb.webp
steps:
  - title: "Open Edit with Prompts"
    text: "Open the image, click the three dots in the bar above it and choose Edit with Prompts. Your image loads under Reference images and the panel picks Nano Banana 2.1."
  - title: "Choose an editing model"
    text: "To change it, click the model name › View model library › Capabilities › Edit with Prompts. This lesson keeps Nano Banana 2.1; if one model changes too much, try the same prompt on another."
  - title: "Write the one change, then what must stay"
    text: "Two separate runs, each with one change and a keep list by name: the first turns the fox into an owl chick, the second turns the whole scene into a winter night. Click Generate for each."
    prompt: |-
      Change only the fox cub into a small fluffy white snowy owl chick with big round amber eyes, sitting in the same place on the robot's palm and looking up at the robot. Keep everything else exactly the same: the robot, its moss and flowers, its glowing eye, its hand, the river, the stones, the trees, the mist, the sunrise light and the framing.

      Change the whole scene to a snowy winter night: a half-frozen river with thin ice, snow on the stones, the banks and the birch trees, falling snowflakes and a deep blue moonlit sky. The robot's eye glows warm amber on the snow. Keep the same robot, its pose, the fox cub on its palm, the camera angle and the framing.
  - title: "Check that nothing else moved"
    text: "Compare each edit with the original at full size. Only what you named may change; the Info panel keeps the prompt, the model and the reference."
  - title: "Retouch one small area"
    text: "Three dots › Retouch, then draw a lasso mask a little larger than the robot's eye; the model switches to Scenario Inpaint (Feather Pixels 32, Padding Ratio 0.1). Describe only the masked area, click Generate, then Select."
    prompt: |-
      The robot's round eye glows a soft bright blue instead of amber, same round shape, same metal rim.
  - title: "Expand to a wider format"
    text: "Click Use as Reference, then the model › View model library › Capabilities › Show all › Expand, and pick Photoroom Expand. Set Output Width 2688 and Output Height 1152 and click Generate."
  - title: "Check the expanded image"
    text: "The original pixels stay and the river and forest continue in the same light. Check the seam where the new part starts for a line or a change of color."
agent_steps:
  - title: "Connect the Scenario MCP"
    text: "Connect your agent and sign in once at [mcp.scenario.com](https://mcp.scenario.com). Then tell it which Scenario team and project to use, so every file lands there."
  - title: "Give the agent the files"
    text: "The image to edit, here the moss-covered robot holding a fox cub at 2048 x 1152. Upload it with upload_asset or give its asset ID. Every edit makes a new asset, so the original stays safe."
  - title: "Brief the agent: two prompt edits"
    prompt: |-
      Use Nano Banana 2.1 (model_google-gemini-nano-banana-2-1) for two separate edits of the robot image. Pass it as referenceImages, an array holding its asset ID, in each run. Read model_schema_get first and keep the original aspect ratio. Quote the cost with a dry run first and wait for my go.

      Edit 1, prompt word for word:
      Change only the fox cub into a small fluffy white snowy owl chick with big round amber eyes, sitting in the same place on the robot's palm and looking up at the robot. Keep everything else exactly the same: the robot, its moss and flowers, its glowing eye, its hand, the river, the stones, the trees, the mist, the sunrise light and the framing.

      Edit 2, prompt word for word:
      Change the whole scene to a snowy winter night: a half-frozen river with thin ice, snow on the stones, the banks and the birch trees, falling snowflakes and a deep blue moonlit sky. The robot's eye glows warm amber on the snow. Keep the same robot, its pose, the fox cub on its palm, the camera angle and the framing.

      Launch both with wait: false and collect them with one jobs_wait, then show each result next to the original.
  - title: "Expand to a banner"
    prompt: |-
      Expand the original robot image with Photoroom Expand (model_photoroom-expand) to 2688 wide and 1152 high, keeping the original pixels. Read model_schema_get for the exact names of its output width and height parameters. Quote the cost with a dry run first and wait for my go.
  - title: "Check and redo"
    text: "Accept when only what you named changed, the robot and fox (or owl) are the same characters, the pose, framing and light held, and the Expand seam carries the water and mist across with no line. Reject a changed color, face or background, or a reframe. Retouch with a painted mask stays in the web app (Step 5)."
    prompt: |-
      For each edit, tell me if anything I didn't name changed: the robot's moss, its eye, the river, the light or the framing. Redo only a failed edit with the same prompt, adding the part that moved to the keep list, or try GPT Image 2.5 Sunburst (model_openai-gpt-image-2-5-sunburst) instead. For the banner, check the seam where the new pixels start.
  - title: "File the results and report"
    prompt: |-
      Create a collection called "Robot edits" with the original, the kept edits and the banner. Report each asset ID with its model and the total cost spent.
updated: 2026-11-11
canonical: https://learn.scenario.com/images/edit-images
---

Last updated: 2026-11-11

# Edit an image without starting over

> Change one element by prompt, repaint a masked area, extend the frame or change the format, and check that nothing else moved.

In short:

- Edit instead of regenerating when most of the image is already right: the rest of the image stays as it is.
- Describe the change in a sentence, paint a mask over the area, or extend the frame: pick the tool that fits the change.
- Every edit makes a new asset, so the original stays safe in your library.


When most of an image is right, edit it instead of starting over. This lesson takes one image, a moss-covered robot holding a fox cub in a misty river, and edits it three ways: **Edit with Prompts** with Nano Banana 2.1 to change one element or the whole scene, **Retouch** to repaint one small area under a mask, and **Expand** with Photoroom Expand to grow it into a wide banner. Each edit makes a new image, so the original stays safe.

:::note As of November 2026
The models, settings and prices in this article are the ones Scenario offered in November 2026. New models and updates arrive often, so check the model library for newer options before you choose.
:::

:::steps Quick version
1. Open the image, click the three dots in the bar above it and choose **Edit with Prompts**, **Retouch** or **Expand**.
2. To change one thing anywhere in the image, use **Edit with Prompts**: write the one change, then list what must stay, and click **Generate**.
3. To change one small area, use **Retouch**: draw a mask around it with the lasso, say what goes there, click **Generate**, then **Select**.
4. To fit a wider or taller format, click **Use as Reference**, pick **Photoroom Expand**, set **Output Width** and **Output Height**, and click **Generate**.
5. Compare the result with the original and check that nothing else moved.
:::

## Before you start

- **You need an image in your library.** Any image works: one you generated in Scenario or one you uploaded.
- **Edit, don't regenerate, when most of the image is right.** A new generation changes everything at once. An edit changes the part you name and keeps the rest, so you don't lose the composition, the light or the character you already approved.
- **Pick the tool by the size of the change:**

| You want to | Use | How you point at the change |
|---|---|---|
| Change one element, the season, the light or the style | **Edit with Prompts** | In words |
| Change one small area and keep every other pixel | **Retouch** | With a mask you draw |
| Fit a wider or taller format | **Expand** | With the output size |

- **Every edit is a new asset.** The original stays in your library, untouched, so you can always go back and try again.

![The original image: a moss-covered robot holding a fox cub in a misty river at sunrise](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/324970d852ae1cc7d7bfb9a3a66391a5be0c93f633b74091ab8f6e812d70f250.webp)

## Step 1: Open Edit with Prompts

Open the image. In the bar above it, click the three dots to open the menu, and choose **Edit with Prompts**.

![The three-dot menu above the image, with Edit with Prompts under the pointer](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/270f000c1d35b9d1545bcc13c8c4fe46963875134a49b4d1667b109f45041d8e.webp)

The same menu holds **Retouch**, **Expand** and **Remove Background**, used later in this lesson and the next. You can also reach Edit with Prompts from the three dots on a thumbnail in your gallery, or from **Create** in the top bar with no image loaded.

Edit with Prompts opens with your image under **Reference images**, ready to edit. It also picks a model for you, here **Nano Banana 2.1**.

![The Edit with Prompts panel with Nano Banana 2.1 picked and the robot image under Reference images](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/5a969d0b325396ede899150b4ce7d662ae3c7843d0cdd6de894b24a8703d1346.webp)

Under the image you can add more references (drag and drop or paste, choose one from your library, or upload one), and the pen icon lets you draw on the image to show the model where to work.

## Step 2: Choose an editing model

To use another model, click the model name, then **View model library**. Under **Capabilities**, tick **Edit with Prompts** to see only the models made for editing. In the film the filter shows 47 models, among them GPT Image 2.5 Flare and Sunburst, Seedream 5.0 Pro, Ideogram 4.5, Gemini 3.0 Pro and Nano Banana 2.1.

![The model library filtered to Edit with Prompts, with Nano Banana 2.1 selected](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/799727a11866d010bf70ffc01a7ad7891b5da878719a8c898c5c19c5c0ed0a40.webp)

This lesson keeps Nano Banana 2.1, so close the library. Editing models differ in how strictly they keep what you didn't mention: some are precise and keep the style, others redraw more freely. When one changes too much, try the same prompt on another.

## Step 3: Write the one change, then what must stay

An editing prompt has two parts:

1. **The one change**, said precisely: what changes, into what, and where it sits.
2. **What must stay**, listed by name: the subject, its pose, the light, the framing. Models keep what you name more reliably than what you leave out.

Here, the fox becomes a small snowy owl chick:

```text
Change only the fox cub into a small fluffy white snowy owl chick with big round amber eyes, sitting in the same place on the robot's palm and looking up at the robot. Keep everything else exactly the same: the robot, its moss and flowers, its glowing eye, its hand, the river, the stones, the trees, the mist, the sunrise light and the framing.
```

![The owl prompt in the Prompt box, with the robot image under Reference images](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/d9f38ab0211cdd9071d94b8ecc5115f2cfb5be2a94b53330b902f9f182271162.webp)

Click **Generate**. The price of the run shows on the button.

The same kind of prompt can change the whole scene too. Make it a snowy winter night, keep the robot and the fox, and click **Generate**:

```text
Change the whole scene to a snowy winter night: a half-frozen river with thin ice, snow on the stones, the banks and the birch trees, falling snowflakes and a deep blue moonlit sky. The robot's eye glows warm amber on the snow. Keep the same robot, its pose, the fox cub on its palm, the camera angle and the framing.
```

Keep each prompt to one or two changes. If you need more, run them one after the other on each new result: small steps are easier to check and to undo.

## Step 4: Check that nothing else moved

Open the result. Here is the owl, sitting in the robot's hand, and nothing else moved: the same moss, the same eye, the same river and light. The **Info** panel keeps the prompt, the model and the reference image the edit was made from.

![The owl edit in the viewer, with Nano Banana 2.1 and the prompt in the Info panel](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/cbeadd201a89fc407d2ac0a698dd3191d546ebe13020b57b365224a1289d5aed.webp)

And here is the same moment on a frozen winter night. The robot, its pose, the fox and the framing all held.

![The winter night edit in the viewer: snow on the robot, a moonlit sky, the fox still on its palm](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/c5a7c76fb07bde864e1b6c417e5291dd4712424676b43890bf3ba4b7feb5ed89.webp)

Compare each edit with the original side by side, at full size:

:::compare
![The original: the fox cub on the robot's palm at sunrise](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/324970d852ae1cc7d7bfb9a3a66391a5be0c93f633b74091ab8f6e812d70f250.webp "Original")
![The edit: a snowy owl chick in the same place, everything else unchanged](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/01f2c90c418ab11f524d025efa9dd9a019e62d6cb5970255791faebf78a01b76.webp "Edit with Prompts")
:::

:::check
**Accept when:** only what you named changed, the subject is the same character, the pose and framing match, and the light falls the same way on everything you kept.
**Reject when:** something you didn't name changed (a color, a face, the background), the new element floats or has the wrong light, or the image was reframed. Name the part that moved in the keep list and generate again, or try another editing model.
:::

## Step 5: Retouch one small area

For a small area you can point at, open the three-dot menu and choose **Retouch**. It is Scenario's inpainting tool: it repaints only the area you mask.

![The three-dot menu with Retouch under the pointer](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/c42e5bf0dd45009401cee38572388f91d420f436b251bc0af784ac6fa380711a.webp)

The image opens on a canvas. On the left are the **Model**, **Prompt**, **Reference** and **Settings**; the vertical toolbar next to the image holds the selection tools (brush, lasso and smart selection), the sketch and erase tools, import, undo and redo.

![The Retouch canvas with the image loaded, the generation panel on the left and the toolbar beside the image](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/5ef2ddb71fa92b82c05b5266411ab68b07ab7cc88d001c5e72d60382c651ab09.webp)

1. Pick the lasso and draw around the robot's eye. Only the masked area can change. The mask shows over the image and in the panel under **Reference & Mask**, and the model switches to **Scenario Inpaint**.
2. Say what goes there. The prompt describes only the masked area, not the whole scene:

```text
The robot's round eye glows a soft bright blue instead of amber, same round shape, same metal rim.
```

The film leaves the two **Settings** of Scenario Inpaint, **Feather Pixels** and **Padding Ratio**, at 32 and 0.1. Click **Generate**.

![The lasso mask over the robot's eye, Scenario Inpaint picked and the blue-glow prompt in the panel](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/e2c655edfcf514ee1768d5d43b8ee17d05d6ff15f596b147fc3f298f9e60075e.webp)

3. The result appears on the canvas, with its thumbnail at the top left. The eye glows blue, and the rest is untouched. Click **Select** to keep it, or **Cancel** to drop it and try again.

![The Retouch result: the robot's eye glows blue, with Select, Deselect all and Cancel on the left](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/f536d5d76343d799976f71b433a7eaf9d591cc5e803427143f07de021741510f.webp)

:::compare
![The original: the robot's eye glows amber](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/324970d852ae1cc7d7bfb9a3a66391a5be0c93f633b74091ab8f6e812d70f250.webp "Original")
![After Retouch: only the eye changed, it now glows blue](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/b364aa0782d9f990ef0eed2a50afe221b2c9a1789a757220de007547b467a06a.webp "Retouch")
:::

Draw the mask a little larger than the thing you change, so the new pixels blend into the old ones. The area around the mask guides the result, which is why the new eye keeps the same light as the rest of the robot.

## Step 6: Expand to a wider format

To fit a wider format, open the image and click **Use as Reference** in the bar above it. Your image goes into the panel as the reference, with the current model, here **GPT Image 2.5 Sunburst**.

![The image loaded under Reference Images after Use as Reference, with GPT Image 2.5 Sunburst picked](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/4ea90bc47763a61e619f4e784918700321be07c9d9396f954fdc741b92a95ea8.webp)

1. Click the model, then **View model library**.
2. Under **Capabilities**, click **Show all** and tick **Expand**. Three models do it: **Smart Reframe**, **Photoroom Expand** and **Reframe**. Pick **Photoroom Expand**.

![The model library filtered to Expand: Smart Reframe, Photoroom Expand and Reframe](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/63a9efd290038b5d33532013beca65ec25ead35c3db4e5f43865c37f09edfc6b.webp)

3. Photoroom Expand puts your image under **Image** and has three settings: **Output Width**, **Output Height** and **Seed**. Set the **Output Width** to 2688 and the **Output Height** to 1152, a wide banner, and click **Generate**.

![Photoroom Expand with the image loaded and the Output Width, Output Height and Seed settings](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/4974fee4f2baaf4e0abafff622c836833d530e39f4435079dd49b7ff145088fc.webp)

The original is 2048 × 1152, so the height stays the same and the width grows by 640 pixels. The model keeps the original pixels and draws new ones in the space around them. The output must be at least as large as the original on both sides: Expand never shrinks an image.

## Step 7: Check the expanded image

Your original is untouched, and the river and the forest carry on in the same light. The **Info** panel shows **Photoroom Expand**, the **Output Width** and **Output Height**, and the seed it used.

![The 2688 × 1152 result in the viewer, with Photoroom Expand and its settings in the Info panel](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/cf244b7a1dccf7ec534af8ed3a2a50c5afa79c30497dbcdcfcaadb0712b410cb.webp)

![The expanded banner: the river and forest extend on the left in the same sunrise light](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/1f5fbd575ba35e993b2d74a33b1eb957f4d4d98c5e30f74cd12b3b6a279b6476.webp)

Look at the seam where the old image ends and the new part starts. The water, the mist and the shadows should carry across it without a line or a change of color.

## Photoroom Expand or Smart Reframe

| Model | What it does | Settings |
|---|---|---|
| **Photoroom Expand** | Keeps the original as it is and draws new background around it. Good for turning a square shot into a banner or a post into a story. | Output Width, Output Height, Seed |
| **Smart Reframe** | Recomposes the frame to an exact size instead of only adding to the sides, and keeps the style, the subject, on-image text and the camera angle. | Width and Height (with presets), an optional prompt, Text Density, Image Count, Thinking Level |

The filter also lists **Reframe**. Use Photoroom Expand when the original must stay pixel for pixel. Use Smart Reframe when the layout can move, for example to fit text or a logo into a new format.

## Common problems

| What you see | What to do |
|---|---|
| The edit changed things you didn't ask for | List them in the keep part of the prompt by name, or try another editing model |
| The new element looks pasted on, with the wrong light | Say where the light comes from and ask the model to match it, or use Retouch with a slightly larger mask |
| The edited image is smaller than the original | Some editing models return their own size: check **Dimensions** in the Info panel, and set **Aspect Ratio** or the resolution in **Settings** if the model has them |
| Retouch left a visible edge around the mask | Draw the mask a little wider so it takes in some of the surroundings, and generate again |
| Retouch changed nothing | Make the prompt say what is different ("blue instead of amber"), not only what is there |
| A seam shows where Expand added pixels | Generate again with another seed, or try Smart Reframe |
| Expand won't run | The output must be at least as large as the original on both sides |

## What's next

To cut a subject out or put it somewhere new, go to [Remove and replace backgrounds](https://learn.scenario.com/images/remove-and-replace-backgrounds.md). When an edited image is approved, [upscale it](https://learn.scenario.com/images/upscale-and-enhance-images.md) as the last step.

Source: https://learn.scenario.com/images/edit-images