nano-banana

Generate and edit AI images via Gemini's REST API using a Python CLI.

Updated Feb 5, 2026
One-click install
npx skills add https://github.com/difflabai/marketplace --skill nano-banana-difflabai
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: nano-banana
Source: https://github.com/difflabai/marketplace/tree/main/plugins/nano-banana/skills/nano-banana
Command: npx skills add https://github.com/difflabai/marketplace --skill nano-banana-difflabai

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Nano Banana provides a straightforward path to create and refine AI-generated images using Gemini's REST API, without the need for MCP servers or npm installs.

Core Features & Use Cases

  • Generate new AI images from natural language prompts using Gemini models.
  • Edit existing images with textual instructions and maintain image lineage via an on-disk history log.
  • Iterate on image series by continuing edits and referencing prior generations.

Quick Start

Invoke the bundled Python CLI (scripts/generate.py) to generate or edit an image with your prompt and instructions.

Frequently Asked Questions about nano-banana

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate AI images using the Gemini API without an MCP server?

You can generate AI images using the Gemini API without an MCP server by running a self-contained Python CLI. This script uses your GEMINI_API_KEY to process natural language prompts and outputs the final image file path directly.

Can I edit existing images and keep an on-disk history of AI image edits?

Yes, you can edit existing images and maintain an on-disk history log. The Python CLI tracks image lineage, allowing you to reference prior generations and continue iterating on image series seamlessly.

Do I need npm or an MCP server to use Gemini for image generation and editing?

No, you do not need npm or an MCP server to use Gemini for image generation and editing. The tool operates as a standalone Python CLI script, requiring only the GEMINI_API_KEY to function.

How do I iterate on an image series by continuing edits with the Gemini REST API?

To iterate on an image series, you continue edits by passing textual instructions to the Python CLI. It references the on-disk history log of prior generations to maintain image lineage and build upon previous outputs.

What is the best way to generate AI art from text prompts via the Gemini REST API?

The best way to generate AI art from text prompts via the Gemini REST API is using a direct Python CLI. It bypasses complex server setups, writes history to disk, and prints the generated image path for easy access.

Are there limitations when using a Python CLI for Gemini image generation?

The primary limitation of using this Python CLI for Gemini image generation is the dependency on the GEMINI_API_KEY. It operates entirely as a local script without a server interface, managing history directly on disk.