ComfyUI Extension: ComfyUI_Gemini_3_Pro_API_node

Authored by Seb-Lis

Created

Updated

0 stars

Run ComfyUI workflows without the setup

No installs, no CUDA version roulette, no GPU sitting idle on your bill. Bring a workflow and run it in the browser.

Custom node for ComfyUI that generates images using Google's Gemini API directly within ComfyUI, supporting resolutions up to 4K.

Looking for a different extension?

Custom Nodes (1)

README

ComfyUI_Gemini_3_Pro_API_node

Gemini_3_Pro_API_node for ComfyUI

GEMINI IMAGE GENERATION NODE - INSTALLATION GUIDE

===================================================================

This custom node allows you to generate images using Google's Gemini API

directly within ComfyUI, supporting resolutions up to 4K.


QUICK START - COPY & PASTE THESE COMMANDS


1. Navigate to your ComfyUI directory in terminal/command prompt

2. For Windows (PowerShell or CMD):

 

  cd C:\Users\[YourUsername]\Documents\ComfyUI

  .venv\Scripts\activate

  pip install --upgrade google-genai

 

3. For Linux/Mac:

 

  cd ~/ComfyUI

  source venv/bin/activate

  pip install --upgrade google-genai

4. Verify installation (optional):

 

  pip show google-genai

5. Set API key as environment variable (recommended):

 

  Windows (PowerShell):

  $env:GOOGLE_API_KEY="YOUR_API_KEY_HERE"

 

  Windows (CMD):

  set GOOGLE_API_KEY=YOUR_API_KEY_HERE

 

  Linux/Mac:

  export GOOGLE_API_KEY="YOUR_API_KEY_HERE"

6. Restart ComfyUI


REQUIREMENTS


- ComfyUI installed and working

- Python 3.8 or higher

- Google AI Studio API key (free tier available)

- Internet connection


INSTALLATION STEPS


1. INSTALL THE NODE

 

  Copy the "Google Gemini API" folder to your ComfyUI custom nodes directory:

 

  Windows:

  C:\Users\[YourUsername]\Documents\ComfyUI\custom_nodes\

 

  Linux/Mac:

  ~/ComfyUI/custom_nodes/

2. INSTALL PYTHON DEPENDENCIES

 

  Open a terminal/command prompt and navigate to your ComfyUI directory, then run:

 

  Windows (PowerShell):

  .\venv\Scripts\activate

  pip install --upgrade google-genai

 

  Linux/Mac:

  source venv/bin/activate

  pip install --upgrade google-genai

 

  IMPORTANT: You need google-genai version 1.51.0 or higher for 4K support.

 

  To verify the version:

  pip show google-genai

3. GET YOUR GOOGLE API KEY

 

  a) Go to: https://aistudio.google.com/app/apikey

  b) Sign in with your Google account

  c) Click "Create API Key"

  d) Copy the generated key (starts with "AIza...")

 

  BILLING NOTE: Image generation requires billing to be enabled.

  - Free tier has ZERO quota for image generation

  - You MUST enable billing at: https://console.cloud.google.com/billing

  - Set up budget alerts to avoid unexpected charges

4. CONFIGURE THE NODE

 

  Option A: In ComfyUI Node

  - Paste your API key directly into the "api_key" field

 

  Option B: Environment Variable (Recommended for security)

  - Set GOOGLE_API_KEY environment variable

  - The node will automatically use it if the field says "INSERT_API_KEY_HERE"

 

  Windows (PowerShell):

  $env:GOOGLE_API_KEY="YOUR_API_KEY_HERE"

 

  Linux/Mac:

  export GOOGLE_API_KEY="YOUR_API_KEY_HERE"

5. RESTART COMFYUI

 

  After installing dependencies, restart ComfyUI completely.


NODE FEATURES


- Models:

  * gemini-3-pro-image-preview (recommended, supports 4K)

  * gemini-2.5-flash-image

- Resolutions:

  * 1K: ~1024 pixels

  * 2K: ~2048 pixels

  * 4K: ~4096 pixels (e.g., 5504x3072 for 16:9)

- Aspect Ratios:

  * 1:1, 16:9, 4:3, 3:4, 9:16, 2:3, 3:2, 4:5, 5:4, 21:9


BUDGET PROTECTION


To avoid unexpected charges, set up a budget alert:

1. Go to: https://console.cloud.google.com/billing

2. Select your billing account

3. Click "Budgets & alerts" (left menu)

4. Click "Create Budget"

5. Set monthly limit (e.g., $5 or $10)

6. Enable email alerts at 50%, 90%, 100%

Image generation costs vary by resolution. 4K images cost more than 1K.


TROUBLESHOOTING


ERROR: "RESOURCE_EXHAUSTED" or "429" error

SOLUTION: Enable billing for your Google Cloud project. Free tier has

  zero quota for image generation.

ERROR: "GenerateContentResponse object has no attribute 'parts'"

SOLUTION: Outdated SDK. Run: pip install --upgrade google-genai

ERROR: "Image object has no attribute 'mode'"

SOLUTION: Update GeminiImageGen.py to latest version from this folder.

ERROR: "ImageConfig not found"

SOLUTION: Upgrade google-genai to 1.51.0+: pip install --upgrade google-genai

ERROR: Images always 1K regardless of setting

SOLUTION: Ensure google-genai is version 1.51.0 or higher.


SUPPORT


For issues with:

- This node: Check the GeminiImageGen.py file for updates

- Gemini API: https://ai.google.dev/gemini-api/docs/image-generation

- Billing: https://console.cloud.google.com/billing

===================================================================

Last Updated: 2025-11-20

===================================================================

Run ComfyUI workflows without the setup

No installs, no CUDA version roulette, no GPU sitting idle on your bill. Bring a workflow and run it in the browser.

Learn more