AI Combine FeatureUpdated 7 hours ago
What Is AI Combine?
AI Combine is an AI feature in the FMX Workflow Builder that lets you generate AI visuals using both text prompts and reference images. This way, the AI doesn’t just guess what you want — it sees example visuals and uses them to guide the output, making results more predictable and aligned with brand or event requirements.
How it works:
You provide reference images
You write a descriptive prompt
AI Combine uses both to generate a precise result that stays true to your reference assets
This is especially helpful when you need consistent, repeatable, and brand‑aligned outputs across many guests or sessions.
Here is a workflow that uses AI Combine for reference:
The AI combine feature goes after the take photo feature, just like other AI workflows.
AI Models
When it comes to choosing your AI Model, it really boils down to 3 things.
The number of people who are expected to use the booth at the event
What AI model are you familiar with using
Time allotment per AI photo generation
The more detailed the AI Model you choose, the longer the AI will take to produce the final output.
Depending on the quality that you want to produce and the time allotted per photo, you can play with the different models that we have.
You can also play with how the workflow is set up for your upcoming event.
Prompting
When it comes to prompting, we highly suggest that you first choose your AI model.
After you have chosen your AI model, you can get help from ChatGPT with creating a prompt.
Here is an example that you can use when asking Chat GPT to create a prompt:
"Can you create me a prompt for *AI model*, e.g., Nano Banana Pro, My reference images are *reference images* *scenario that you desire to achieve* ex (make the person hold this, wear this, and pose with reference image)."
Reference Images
Add your reference images to the cloud by clicking on “ADD REFERENCE IMAGE”.
Have your reference images saved locally and upload them to your assets on the cloud.
Once you have chosen an AI model, input a prompt, and uploaded reference images, you can proceed to test your AI Combine Workflow!
Here are some examples:
