Building a No-Code n8n Photoshop AI Agent with Nano Banana

The way we create and manage digital content is being reshaped by AI. Instead of relying on heavy software and manual processes, we can now build intelligent, no-code agents that handle creative tasks directly from chat apps. One powerful example is the n8n Photoshop AI Agent, which integrates Google Drive, Telegram, and Fal.AI’s NanoBanana model to generate and edit photorealistic images—entirely through natural language prompts.

This project shows how automation and AI can merge: a chat-based assistant that organizes your media, edits and combines photos, and delivers final outputs without ever touching Photoshop.

Core Capabilities of the Agent

Download workflow (n8n): https://romhub.io/n8n/Photoshop_AI_Agent

The Nano Photoshop Agent connects a Telegram bot with Google Drive and two custom n8n workflows—Combine Images and Edit Image—to act like a personal design studio . Its main functions are:

1. File Management via Google Drive

• Photo Uploads: Any photo sent to Telegram is uploaded to Drive automatically.
• Naming on the Fly: The agent immediately asks, “What would you like me to name that photo in your Google Drive?” and then renames the file using the Change Name tool .
• Search Tools: Two dedicated Google Drive queries allow the agent to fetch either raw user uploads or past AI-generated images.

2. Image Generation Workflows

• Combine Images Workflow:
  • Accepts a text prompt plus two Drive file IDs.
  • Downloads the files, uploads them to ImgBB to get public URLs, and passes them to Fal.AI’s NanoBanana model .
  • Polls until the result is ready, then uploads the new composite back into Drive with the chosen title.
• Edit Image Workflow:
  • Works nearly the same way, but with a single image ID and a new creative prompt .
  • Perfect for tasks like “Place this product in front of the Eiffel Tower” or “Add a beach background.”

Both workflows are modular, meaning they can be reused in any other agent or project.

How It Works in Practice

Here’s what interacting with the agent looks like:

1. Upload & Name You send a selfie to Telegram. The agent uploads it to Drive and asks for a name. You reply “Nate,” and now the file is neatly stored and indexed.
2. Combine You send a bag-of-granola image, name it “granola,” and then prompt: “Combine Nate and granola so it looks like the man is holding the bag on a mountain hike.” The agent fetches both IDs, runs the Combine workflow, and delivers a photorealistic composite back in chat.
3. Edit Next, you ask: “Make an ad where the granola is held in front of the Eiffel Tower.” The Edit workflow takes the “granola” image, runs NanoBanana with the new context, and sends back the final visual.

All of this is done conversationally, without needing to open Photoshop.

Inside the Workflows

The magic lies in how n8n nodes are orchestrated:

• Telegram Trigger + Switch: Detects if input is text or photo .
• Google Drive Nodes: Handle uploads, downloads, and updates.
• ImgBB Uploads: Convert binary files to public URLs for Fal.AI.
• Fal.AI NanoBanana: Generates new images based on prompt + URLs.
• Polling System: Waits until generation finishes, then retrieves the result.
• Memory & LLM: The agent uses GPT-5 mini with Claude Sonnet 3.5 as backup for reasoning, keeping chat context with session memory .

Cost, Expansion, and Future Use

Running on Fal.AI, NanoBanana costs just a few cents per image—affordable for rapid prototyping and ad variations. The modular design means you can add:

• Dedicated prompt-optimization agents for sharper outputs.
• Logging into Google Sheets for monitoring usage and costs.
• Video generation tools to extend from stills to motion graphics.

The Nano Photoshop Agent proves that with n8n, you can build a fully functional AI-powered creative assistant—no code required. It takes the complexity out of Photoshop-level tasks, letting you control everything from a chat app.

If your workflow involves content creation, marketing assets, or rapid prototyping, this setup offers a streamlined, cost-effective way to produce high-quality visuals. With Google Drive as storage, Telegram as the interface, and NanoBanana as the creative engine, you’ve got a full AI design studio in your pocket.