image-gen-cli

A multi-provider image generation CLI supporting Google Gemini, OpenAI, BFL FLUX, and Poe image models.

Features

• Generate images from text prompts
• Edit existing images with text instructions
• Multiple provider support (Gemini, OpenAI, BFL, Poe)
• Configurable aspect ratios and output sizes
• Generate multiple images in a single request
• Includes an Agent Skill for AI agent integration

Installation

Requirements: .NET 10.0 SDK

Install as Global Tool (Recommended)

# Clone the repository
git clone https://github.com/PederHP/image-gen-cli.git
cd image-gen-cli

# Build and install
dotnet pack src/ImageGenCli.csproj -c Release -o ./nupkg
dotnet tool install --global --add-source ./nupkg ImageGenCli

Or use the install script:

./image-gen/scripts/install.sh      # Linux/macOS
./image-gen/scripts/install.ps1     # Windows PowerShell

Run from Source

dotnet run --project src/ImageGenCli.csproj -- "Your prompt here"

Configuration

Set at least one API key as an environment variable:

# For Gemini (default provider)
export GEMINI_API_KEY=your-api-key

# For OpenAI
export OPENAI_API_KEY=your-api-key

# For BFL (FLUX)
export BFL_API_KEY=your-api-key

# For Poe
export POE_API_KEY=your-api-key

Or pass it directly with --api-key.

Usage

# Basic image generation (uses Gemini by default)
image-gen "A sunset over mountains with a lake reflection"

# Use OpenAI provider
image-gen -p openai "A futuristic city skyline at night"

# Use BFL FLUX provider
image-gen -p bfl "A cyberpunk street scene with neon signs"

# Use Poe provider (access to many models)
image-gen -p poe "A watercolor landscape"
image-gen -p poe -m Imagen-4 "A photorealistic portrait"

# Specify aspect ratio
image-gen -a 16:9 "Desktop wallpaper, abstract geometric art"

# Generate multiple images
image-gen -n 4 "Logo concepts for a coffee shop"

# Edit an existing image
image-gen -i photo.jpg "Remove the background and add a beach scene"

# Use multiple reference images
image-gen -i style.png -i content.jpg "Apply the style to the content"

# Save to specific directory
image-gen -o ./output "Product photography, ceramic mug"

# High quality with Gemini Pro
image-gen -m gemini-3-pro-image-preview -r 2K "Detailed architectural blueprint"

# List available models for a provider
image-gen --list-models -p poe

Options

Option	Short	Default	Description
--provider	-p	gemini	Provider: gemini, openai, bfl, or poe
--model	-m	auto	Model name (use --list-models to see options)
--images	-i	none	Reference image paths (can specify multiple)
--aspect-ratio	-a	1:1	Aspect ratio (1:1, 2:3, 3:2, 3:4, 4:3, 9:16, 16:9, etc.)
--resolution	-r	1K	Resolution: 1K, 2K, 4K (Gemini Pro, BFL, some Poe models)
--quality	-q	none	Quality: low, medium, high (OpenAI, Poe)
--temperature	-t	1.0	Temperature 0.0-2.0 (Gemini only)
--system-prompt	-s	none	System instruction (Gemini only)
--samples	-n	1	Number of images (1-4 Gemini, 1-10 others)
--output	-o	current dir	Output directory
--api-key	-k	env var	API key override
--list-models	-l	-	List available models for the provider

Provider Comparison

Feature	Gemini	OpenAI	BFL	Poe
Default model	gemini-2.5-flash-image	gpt-image-1.5	flux-2-pro	GPT-Image-1
System prompts	Yes	No	No	No
Temperature control	Yes	No	No	No
Resolution control	Pro models	No	Yes	Model-dependent
Quality control	No	Yes	No	Yes
Max images per request	4	10	10	10
Max reference images	N/A	N/A	8	N/A

Note: Using unsupported options with a provider will result in an error. For example, --temperature with OpenAI will fail.

Poe Provider Privacy Warning

Important: Images generated through the Poe provider are returned as URLs that are publicly accessible without authentication. While the URLs are not easily guessable, anyone with the link can view the image. Do not use the Poe provider for generating sensitive or private content if URL confidentiality is a concern.

Output

Generated images are saved with the naming pattern:

• Single image: {provider}-{timestamp}.{ext}
• Multiple images: {provider}-{timestamp}-{n}.{ext}

For AI Agents

This project includes an Agent Skill definition at image-gen/SKILL.md. The skill provides structured instructions that AI agents can load on-demand to use this CLI effectively.

See also:

• AGENTS.md - General AI agent instructions
• CLAUDE.md - Claude Code specific guidance
• .github/copilot-instructions.md - GitHub Copilot instructions

Development

# Build
dotnet build src/ImageGenCli.csproj

# Run tests (output to .tmp/)
dotnet run --project src/ImageGenCli.csproj -- -o .tmp "test prompt"

Adding a New Provider

1. Create NewProviderClient.cs implementing IImageGenerationClient
2. Update Program.cs with provider options and validation
3. Update documentation files

License

MIT