Skip to main content

Generate images using Claude and Photoroom

How to use Claude to generate images using Photoroom

C
Written by Craig Wright

After adding the Photoroom connector to Claude, you can use natural language prompts to generate images directly in your Claude conversations.

Screenshot of a Claude prompt using the Photoroom MCP to remove the background from an image.

Note: The Photoroom connector uses Photoroom API credits, not AI credits.

Before you begin

Make sure you have:

Note: The Photoroom connector does not use AI credits, even if AI tools are used. API credits are used instead.
​

Steps

To generate images using natural language with Claude and Photoroom:

  1. Log in to Photoroom and select the Space that has API enabled.

  2. Log in to Claude.

  3. Make sure you select the profile that has the Photoroom connector set up.

  4. Start a new chat.

  5. Turn the Photoroom connector on:
    ​

    1. Select + at the bottom left of the chat interface.

    2. Select Connectors.

    3. Find Photoroom and toggle it on.
      ​

  6. Enter a prompt that describes what you want Photoroom to create:
    ​

    • Begin your prompt with "Use the Photoroom MCP to ...". See the example at the end of this section.
      ​

    • Describe what you want Photoroom to do. Include any parameters that you want to apply, for example, the name of the Virtual Model you want to use or the style of background.
      ​

    • Use + to add source images for the AI to use as a starting point. For example, add a product photo.
      ​

  7. Send your prompt and wait for Claude and Photoroom to generate your images.

  8. Download the images.

Tip: You can build more complex prompts that use the Photoroom connector as part of a longer chain involving other products. That's how you can use Claude, Photoroom, and other apps to create efficient, automated workflows.

Example prompt

In the video above (see Steps), we used the following prompt to generate 3 images:

Prompt:

Use Photoroom MCP to create a virtual model image showing Alex in an urban stairwell. Also create a ghost mannequin image of the T-shirt and a flat lay image with a white background.

Plus we added a photo of a T-shirt as the source image (starting point).

Note how we are able to:

  • Use natural language to describe what we want

  • Specify the parameters we wanted, such as the particular virtual model to use

  • Generate 3 images at the same time in Claude

Did this answer your question?