> For the complete documentation index, see [llms.txt](https://helpcenter.channable.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://helpcenter.channable.com/manage-improve-product-data/generate-with-ai/ai-generated-images/ai-generated-images.md).

# How to create and use AI generated images

Generate, preview, review, and approve AI product images with a Gemini or OpenAI connection.

> #### **Who can use this feature**
>
> You need the **Core Pro** plan in your [Channable subscription](/account-billing/manage-your-subscription/channables-pricing-plans.md) to use this feature.

### <i class="fa-memo-circle-check">:memo-circle-check:</i> Before you start

**You'll need**

* A **Core Pro** plan in your [Channable subscription](/account-billing/manage-your-subscription/channables-pricing-plans.md).
* An image-type project field containing the source images.
* A [Gemini](/manage-improve-product-data/generate-with-ai/set-up-your-ai-generation-connection/how-to-create-a-gemini-connection-for-ai-generation.md) or [OpenAI](/manage-improve-product-data/generate-with-ai/set-up-your-ai-generation-connection/how-to-create-an-openai-connection-for-ai-generation.md) connection, or an available **Trial** connection.

{% hint style="info" %}

#### Trial connection

Use the **Trial** connection to test image generation.

* The **Trial** option uses Channable's own **Gemini** connection. You don't pay for this.
* You can generate up to 5 images with the **Trial** connection. Once you've generated 5 images, you'll need to connect to Gemini or OpenAI to generate more.
* After you connect Gemini or OpenAI, the trial connection is unavailable.
  {% endhint %}

**Good to know**

* The generator uses a source image. It cannot create a product image without one.
* Generated images do not appear in the output field until you [approve](#approve-or-regenerate-each-image) them.
  * When you approve an image, it is saved to the output field you select in step 1.5. Choose an empty field to keep the source image. Choose the source field to replace the source image.
* Read [AI generation usage and costs](/manage-improve-product-data/generate-with-ai/ai-generation-usage-and-costs.md) to understand token usage, pricing, and data handling.
* Channable uses Gemini `Gemini-3.1-flash-image` and OpenAI `gpt-image-2.5 Sunburst`.

***

{% stepper %}
{% step %}

### Begin the setup

1. In your project, go to **Generate with AI**.
2. Click **Generate with AI**.
3. From the content type options, select **Images**.
4. Click **Continue**.
   {% endstep %}

{% step %}

### Set the source image and output field

1. In the **Name** field, enter a descriptive name.
2. At **AI connection**, select **Trial**, **Gemini**, or **OpenAI**.
3. If you select **Gemini** or **OpenAI**, select an existing connection or [create a new Gemini](/manage-improve-product-data/generate-with-ai/set-up-your-ai-generation-connection/how-to-create-a-gemini-connection-for-ai-generation.md) or [OpenAI connection](/manage-improve-product-data/generate-with-ai/set-up-your-ai-generation-connection/how-to-create-an-openai-connection-for-ai-generation.md).

{% hint style="info" %}

#### How to choose an AI connection for images

* Gemini is twice as expensive as OpenAI.
* Gemini is typically slightly faster and more accurate than OpenAI, but the difference is minimal.
  {% endhint %}

4. At **Image to use**, select the image-type field containing your source image, such as `image_link`.

{% hint style="warning" %}

#### No source image field?

This error typically happens because the field has not been set correctly to the **Image** type.

To fix this, In **Project fields**, locate the field and set its type to **Image**. Return to the generator and select it.
{% endhint %}

5. At **Field to save the optimized image**, choose where to save approved images:
   * Select the source image field to replace the source image.
   * Create a new image-type field in the flow, such as `image_ai`, and select it to keep the source image.
6. Click **Save and continue**.
   {% endstep %}

{% step %}

### Filter items for image generation

{% hint style="info" icon="lightbulb" %}

#### Tip

Start with one product category and refine the prompt for that category. Create separate prompts for other categories instead of using one prompt for all items.
{% endhint %}

1. In **Select items**, set the filter for products that need generated images.
2. To select every item, leave the condition as **If** `all`.
3. To select a group, replace `all` with a project field and enter a condition. For example, set **If** `title` contains `dress`.<br>

   <figure><img src="https://501928696-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fu8whzIfip72Kvn3M1bsc%2Fuploads%2F93QqfTQYBpjveHHugbWa%2FScreenshot%202026-09-23%20at%2015.21.47.png?alt=media&amp;token=9bde2584-b1b9-4049-b7d4-121e8a6aa0d9" alt=""><figcaption></figcaption></figure>
4. Click **Apply** to update the selected items.
5. Click **Save and continue**.

{% hint style="warning" %}

#### Important

Review the selected items before generating images. Items without a source image cannot produce a useful product-based result, and will return an `Input image not found` error.
{% endhint %}
{% endstep %}

{% step %}

### Create and review image previews

1. In **Prompting**, review the example prompt and describe the image you want to create.
2. At **Image ratio**, select the ratio that suits your destination.
3. Check the estimated preview cost.<br>

   <div align="left"><figure><img src="https://501928696-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fu8whzIfip72Kvn3M1bsc%2Fuploads%2FCYejDgphgu3C2XSxmrSA%2FScreenshot%202026-09-23%20at%2015.38.46.png?alt=media&amp;token=83b899cb-f012-4d91-914c-476f80cce238" alt="" width="375"><figcaption></figcaption></figure></div>
4. Click **Generate previews**.
5. Review each preview and refine the prompt if needed.

<figure><img src="https://501928696-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fu8whzIfip72Kvn3M1bsc%2Fuploads%2FhLVXZ9kwe9uLtHhas4MY%2FScreenshot%202026-09-23%20at%2015.27.43.png?alt=media&amp;token=c425dac5-5030-4e1a-a5ce-85b9f2b514a3" alt=""><figcaption></figcaption></figure>

Click **Compare to original** to view the source and generated images together. Click **Large preview** to inspect an image more closely. When multiple previews are available, click **Next** to review each item.<br>

{% hint style="warning" %}

#### Important

Each new preview generation adds a cost.
{% endhint %}
{% endstep %}

{% step %}

### Generate images from approved previews

1. Check the estimated generation cost. If you’re using the [**Trial**](#before-you-start) connection, you can generate up to 5 images for free.
2. Click **Generate images**.
3. Wait for the images to finish generating.

Use the **Generate images** menu to generate more images or regenerate the selection.
{% endstep %}

{% step %}
{% hint style="info" %}

#### Duplicate source images

Channable groups products that use the same source image. It generates one image for each unique source image.

The number of generated images can be lower than the number of items selected. When one generated image applies to multiple products, their IDs appear below the image.
{% endhint %}

### Approve or regenerate each image

We recommend reviewing images one at a time:

1. In **Review**, open an image to compare it with the original.
2. Click **Approve** to save it to the selected field, or click **Regenerate** to create another version.
3. Click **Next** to review the following image.

<figure><img src="https://501928696-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fu8whzIfip72Kvn3M1bsc%2Fuploads%2FAxIO3rFZFfIrRonBXGBe%2FScreenshot%202026-09-23%20at%2015.33.35.png?alt=media&amp;token=46eef28f-7031-46a8-929d-924f1121794a" alt=""><figcaption></figcaption></figure>

To work faster, hover over a thumbnail and select **Approve** or **Regenerate**. Use **Approve all** or **Unapprove all** to manage images in bulk.<br>
{% endstep %}

{% step %}

### \[Optional] Automatically generate images for new matching items

1. In **Review**, click **Automation**.
2. Turn on **Generate content for new items** to generate images for new items matching your filters.

{% hint style="warning" %}

#### Important

Automation only works for active projects.
{% endhint %}

3. If you have tested the results, turn on **Approve new content automatically**.
4. At **Daily generation limit**, enter the maximum images to generate each day.
5. Save the automation settings.

{% hint style="warning" %}

#### Important

Set a daily limit before enabling automation. This controls generation costs when many items are added.
{% endhint %}
{% endstep %}

{% step %}

### Map approved images to your channel

1. In your feed, open the mapping step for the destination image attribute.
2. Map the field containing your approved generated images.
3. Save and run the feed.

The channel receives only images that you approved, unless automatic approval is enabled.
{% endstep %}
{% endstepper %}

### Troubleshooting

<details>

<summary>Channable cannot download the source image</summary>

Check that the source image URL is publicly accessible and returns a supported image file. Then generate the preview again.

</details>


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://helpcenter.channable.com/manage-improve-product-data/generate-with-ai/ai-generated-images/ai-generated-images.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.