Home/HighLevel/Features/Workflows/Setting up AI image generation in HighLevel workflows

Workflows and Automation

Setting up AI image generation in HighLevel workflows

In short

The AI Image Generation action lets you create images inside a HighLevel workflow by providing a text prompt, picking a model and quality, and optionally adding reference images or templates. The resulting picture is saved to Media Storage and returned as a public URL for any downstream step.

AreaWorkflows · Section
Used forProduct mockups, Personalized campaign visuals, On demand images from conversations
PlanPremium action (additional charges per execution)
Setup steps12 steps in 1 task
Last updatedNovember 17, 2026

01Who this is for

  • Marketing managers can produce on brand social graphics without hiring a designer.
  • Sales teams can generate product mockups right away for client proposals.
  • Customer support agents can send visual explanations directly in chat conversations.

02AI image generation actions, explained

The AI Image Generation action appears under AI Actions in the workflow builder. It accepts a prompt, optional reference images, and several settings that influence the visual output. When the workflow runs, the selected AI model creates an image, saves it to Media Storage and provides a public URL.

The action returns three custom values: the image URL, the storage path, and the file name. These variables can be used by any following action, such as email, SMS, or webhook steps. All generated assets are automatically organized in the Media Library for later reuse.

This premium action adds an extra charge per execution, so each generated picture contributes to your usage cost based on model, quality, size, and any added brand context.

03How businesses use AI image generation action

  • Product mockups. When a new product description arrives via webhook, the workflow runs the AI Image Generation action with the description inserted into the prompt. The generated mockup image URL is then attached to an email sent to the vendor for quick visual review.
  • Personalized campaign visuals. A contact tag triggers the action. The workflow builds a prompt that includes the contact’s segment data and optionally a reference image, then sends the generated image URL in an email or MMS to personalize the campaign.
  • On demand images from conversations. When a customer asks for a visual in a chat, the workflow uses the customer's message as the prompt, generates an image, and returns the image URL via SMS/MMS in the same conversation thread.

04The numbers and rules that matter

  • Models available: GPT Image 2, GPT Image 2.5 Flare, GPT Image 2.5 Sunburst, Gemini 3 Pro Image (Nano Banana Pro), Gemini 3.1 Flash Image (Nano Banana 2), Gemini 2.5 Flash Image (Nano Banana).
  • Reference images: Up to five per action.
  • Four templates are available: Product Mockups, Social Post, Promo Flyer, Lifestyle.
  • Additional Settings include Size, Background, Output Format, Design Kit, Brand Voice.
  • Cost per image median: $0.0388 (3.88¢) (per HighLevel's help center, checked November 2026).
  • Cost per image average: $0.0499 (4.99¢) (per HighLevel's help center, checked November 2026).
  • The action is a premium action and incurs additional charges per execution.

05Before you start

  • Configure a Design Kit in your account before selecting it in the action.
  • Set up Brand Voice details in your account before using the Brand Voice option.

06How to set up and use AI image generation action in HighLevel

Configure the AI Image Generation action

  1. Open Automation > Workflows.
  2. Create a new workflow or edit an existing one.
  3. Click the + icon, locate AI Image Generation under AI Actions, and add it.
  4. Enter a clear, descriptive name for the action.
  5. Choose the desired AI model from the available list.
  6. Select the image quality (Auto, High, Medium, or Low).
  7. Type a detailed prompt or insert custom values from contacts or previous actions.
  8. Press Enhance Prompt to expand and refine the description if desired.
  9. Optionally pick a template to pre fill the prompt field.
  10. Optionally add up to five reference images via system upload, media library, or URL.
  11. Optionally adjust additional settings (size, background, output format, Design Kit, Brand Voice).
  12. Save the action.

07What it looks like in HighLevel

08Settings that matter

  • Action Name. A custom label that identifies the step in the workflow and appears in execution logs.
  • Model. Selects which AI image model will generate the picture, influencing style and capabilities.
  • Quality. Controls the level of detail of the output; options are Auto, High, Medium, and Low.
  • Prompt. The text description that tells the model what to create; supports custom values for dynamic content.
  • Templates. Pre written prompts for common image types that can be edited after being loaded.
  • Reference images. Up to five optional images supplied to the model as visual context, added via upload, media library, or URL.
  • Additional Settings. Optional controls for image size, background type, file format, and application of Design Kit or Brand Voice.

09Things to know before you rely on it

  • This is a premium action that adds a charge for each execution.
  • You can attach a maximum of five reference images per action.
  • Quality can only be set to Auto, High, Medium, or Low.
  • Model choices are limited to the list provided in the action.
  • Estimated cost per image ranges from $0.0388 to $0.0499 (per HighLevel's help center, checked November 2026).
  • Design Kit and Brand Voice options appear only after they have been configured in the account.

10How it connects to the rest of HighLevel

Michael's take

Michael Reimer
Michael ReimerFounder of CRM Pros, HighLevel Commander and Certified Admin

I find that being specific in the prompt. Describing subject, style, lighting and composition. Generally produces better results. I start by writing a detailed prompt or using a template, then click Enhance Prompt to let the AI add missing details. From there I’d attach a reference image if I need exact brand elements, set the size and background to match the destination, and finally test the workflow on a few contacts before scaling it up.

Get More InformationOpens on HighLevelPublished November 3, 2026 · Updated November 17, 2026

Common questions

Common questions about AI image generation action.

How much does an AI generated image cost?

Generating an image normally costs between $0.0388 (3.88¢) and $0.0499 (4.99¢). The exact amount depends on the chosen model, quality level, size, and whether Design Kit or Brand Voice is applied (per HighLevel's help center, checked November 2026).

Where is the generated image stored?

Every generated image is automatically saved in Media Storage and is accessible via a publicly available URL. You can also retrieve the storage path or file name as custom values.

Which model should I choose for product visuals?

GPT Image 2.5 Flare works well for creator and social content, product visuals, and quick prototyping. GPT Image 2.5 Sunburst is better for premium campaign creative and polished product imagery.

What does the Quality setting affect?

Quality controls the detail level of the output. Auto lets the system decide, High gives the most detail but takes longer, and Low is faster with less detail.

Can I use dynamic data in the prompt?

Yes. The prompt field supports custom values, allowing you to insert contact fields, webhook payload data, or output from previous actions to create unique images per record.

How do I get a transparent background?

In Additional Settings set Background to Opaque (or leave as Auto) and choose PNG or WebP as the output format, since JPEG does not support transparency.

Next step

Pick the features that matter. Then get the plan that unlocks them.

Explore the features you want, start with the five at the top, and upgrade through my link when you are ready. The bonuses and the team come with it.

The features directory explains every HighLevel feature in plain English. Browse it to see what the platform can do.

Back to the FeaturesSee the full details
⌃

Site preview

Explore the next page

This destination is reserved in the page inventory and is not built yet.