How to Generate AI Images in MindStudio

This tutorial shows how to easily and quickly generate AI images in MindStudio.

MindStudio is the easiest way to create amazing AI-generated images using Google Gemini 3 Image Pro (Nano Banana Pro) and other frontier AI image generation models like:

OpenAI GPT Image 1.5
Black Forest Labs Flux 2 Pro and Flux 2 Dev LoRA
ByteDance Seadream 4.5
Qwen Z Image Turbo
Stability SDXL LoRA and Diffusion 3
Wan 2.6
And several others

Generate images in minutes without requiring a high-powered computer or complex software.

Follow the steps below to build a simple image generation workbench:

1. Create a MindStudio account.

2. Go to “Service Router > Balance” to add funds to your Service Router balance. Add $5 to get started.

MindStudio account balance in Service Router

3. Create a new agent from My Drafts.

4. Click the “Start” block and change the Run Mode to “Workbench”.

Change the workflow run mode to Workbench in MindStudio

5. Click the + button to add a new block and choose “Set Variable”.

Add the Set Variable block in MindStudio

6. Name the variable “prompt”. Let’s generate a cool image of a futuristic workspace in the clouds. Copy and paste the following prompt:

,“A cinematic, ultra-wide photorealistic shot of a sleek, futuristic glass workspace suspended in the clouds. On a floating brushed-aluminum desk, a translucent holographic tablet projects a complex 3D flowchart made of glowing blue light. The text "MindStudio" is clearly written in bold, neon white sans-serif font floating above the chart. The background features a soft golden sunset illuminating the clouds. Macro details, depth of field, highly detailed glass textures.”

Add the prompt to the Set Variable block

7. Click the + again and add a “Generate Image” block.

Add the Generate Image block to the workflow in MindStudio

8. Type “{{prompt}}” in the prompt field to pass your variable into the Image Generation model.

Type "{{prompt}}" into the Prompt setting for the Generate Image block in MindStudio

9. Click “Gemini 2.5 Flash Image” under “Model Settings” to open the Image Model picker.

View all AI image generation models in MindStudio

10. Click the “Use” button for “Gemini 3 Pro Image”.

Select Gemini 2.5 Flash Image to generate the AI image in MindStudio

11. In the configuration panel on the right, scroll down to “Advanced”, then select “Enabled” toggle on the “Multiple Variants” option. This is disabled by default. This allows you to generate multiple variations of this image at once.

Enable to generate multiple image variants from the Advanced settings in the Generate Image block

12. Adjust how many variations to make when you run this workbench. 4 is default. This is a good setting to start with so you do not pay for too many images at once.

13. Press “Run” in the top right (or Cmd/Ctrl+R) to start the workbench.

Run the MindStudio workflow to generate the images

14. When the generations finish, four variants are displayed. Click one to open in full screen.

Multiple image variants are displayed after the run completes in MindStudio

15. Choose the best variant and select it to finish the workbench. Note that images are already hosted on MindStudio CDN and ready to be shared or used in other workbenches. This workbench ends with image generation, but a common next step would be to save the generated image to a variable and pass that as the first frame to a “Generate Video” block to animate the image. Check out other tutorials to learn how to continue!