name: nsfw-ai-image-generator description: "NSFW AI image generation via Atlas Cloud API. Supports uncensored image models including Flux, Seedream, Ideogram, Imagen and more. IMPORTANT: Before calling scripts/generate.py, use Atlas Cloud MCP tools (atlas_list_models, atlas_get_model_info) to find the right model and its parameter schema. Use when user asks to generate NSFW, adult, or uncensored AI images."

NSFW AI Image Generator

Generate uncensored AI images using Atlas Cloud API (30+ image models).

Prerequisites

• Python 3.10+ installed
• uv installed
• ATLAS_CLOUD_API_KEY environment variable set

Available Tools

Usage

Generate an image

uv run scripts/generate.py generate "your prompt" --model flux-pro

List available models

uv run scripts/generate.py models

Popular Model Shortcuts

You can also pass any full model ID with --model.

IMPORTANT for AI Agents

Before calling this script, you MUST first use Atlas Cloud MCP tools:

1. Call atlas_list_models with type="Image" to see all available image models
2. Call atlas_get_model_info with the chosen model ID to get its parameter schema
3. Then call scripts/generate.py generate with the correct --model and parameters

🔞 NSFW AI Image Generator Skill

Generate uncensored AI images with 30+ models — Flux, Seedream, Ideogram, Imagen & more

🌍 Language: English | 中文 | 日本語 | 한국어

A Claude Code skill that generates uncensored NSFW AI images through the Atlas Cloud API. Access 30+ state-of-the-art image generation models — including Flux 2 Pro, Seedream V5, Ideogram V3, Imagen 4 Ultra, and more — without running local GPU hardware or dealing with complex setup.

✨ Why This Skill?

🔓 Uncensored Generation

No content filters, no prompt rewriting, no invisible safety classifiers silently modifying your output. You write the prompt, the model generates exactly what you describe.

🖼️ 30+ Image Models via One API

Switch between Flux, Seedream, Ideogram, Imagen, HunyuanImage, and dozens more without managing separate accounts, API keys, or SDKs. One unified interface for all of them.

💰 From $0.003/Image — Cheapest NSFW AI

Flux Schnell starts at roughly $0.003 per image. No monthly subscriptions, no token packs, no minimum spend. You pay only for what you generate.

🤖 AI Agent Ready

Built as a Claude Code skill. Your AI agent can discover models, check parameter schemas, and generate images autonomously — no human-in-the-loop required for the API calls.

🔒 Privacy First

Images are generated server-side via API and delivered directly to you. No prompts are stored, no generated images are logged, no training on your data.

📦 Installation

Step 1: Add the Skill

npx skills add ristponex/nsfw-ai-image-generator-skill

Step 2: Get Your API Key

Sign up at Atlas Cloud to get a free API key. New accounts include free credits — no credit card required.

Step 3: Set Environment Variable

Mac / Linux:

export ATLAS_CLOUD_API_KEY=your_api_key_here

To persist across sessions, add the line above to your ~/.bashrc, ~/.zshrc, or ~/.profile.

Windows (PowerShell):

$env:ATLAS_CLOUD_API_KEY = "your_api_key_here"

Windows (CMD):

set ATLAS_CLOUD_API_KEY=your_api_key_here

Or use a .env file in the project root:

ATLAS_CLOUD_API_KEY=your_api_key_here

🖼️ Popular Models

These are the most popular models. Atlas Cloud offers 30+ image models total — run uv run scripts/generate.py models or use atlas_list_models to see them all.

Model Selection Guide

🚀 Usage

With Claude Code (Recommended)

Once the skill is installed, simply ask Claude to generate images:

> Generate a fantasy landscape with dramatic lighting using Flux Pro

> Create a photorealistic portrait with Imagen 4 Ultra

> Generate 4 variations of a cyberpunk cityscape with Seedream

Claude will automatically:

1. Look up available models via atlas_list_models
2. Check the model's parameter schema via atlas_get_model_info
3. Call the generation script with the right parameters
4. Save the output images to your local machine

Direct CLI Usage

Generate an image:

uv run scripts/generate.py generate "a surreal underwater city with bioluminescent architecture" --model flux-pro

Use a specific model ID:

uv run scripts/generate.py generate "portrait of a woman in golden hour light" --model black-forest-labs/flux-2-pro

Set image size:

uv run scripts/generate.py generate "wide cinematic landscape" --model flux-pro --size 1920x1080

Set aspect ratio:

uv run scripts/generate.py generate "vertical fashion photo" --model ideogram --aspect-ratio 9:16

Generate multiple images:

uv run scripts/generate.py generate "abstract art" --model flux-schnell --count 4

Custom output directory:

uv run scripts/generate.py generate "your prompt" --model flux-pro -o ./my-images

Pass extra model-specific parameters:

uv run scripts/generate.py generate "your prompt" --model flux-pro --extra '{"guidance_scale": 7.5}'

List all available models:

uv run scripts/generate.py models

Model Shortcuts

For convenience, you can use short names instead of full model IDs:

💰 Price Comparison vs Alternatives

Why Atlas Cloud Wins for NSFW

• No subscription lock-in — Pay $0.003–$0.08 per image, nothing more
• No GPU required — All rendering happens on cloud infrastructure
• No censorship — Generate what you want without prompt policing
• Model variety — 30+ models through one API key, switch anytime
• No waitlists — Sign up and start generating immediately
• Free credits on signup — Test before spending anything
• API-first design — Easy to integrate into any workflow or application

🔧 Local Alternatives

If you prefer running models on your own GPU, here are popular open-source options for uncensored image generation:

Flux Uncensored V2 LoRA

• Source: enhanceaiteam/Flux-Uncensored-V2 on HuggingFace
• Requirements: 24GB+ VRAM (RTX 4090 or A100)
• Pros: High quality, based on Flux architecture
• Cons: Requires significant VRAM, LoRA fine-tune only

CHROMA

• Source: lodestones/Chroma
• License: Apache 2.0
• Requirements: 16GB+ VRAM
• Pros: Open license, good general quality
• Cons: Less mature than Flux-based models

Z-Image-Turbo

• Source: Alibaba (open source, 6B parameters)
• Requirements: 12GB+ VRAM
• Pros: Fast inference, smaller model size
• Cons: Less photorealistic than larger models

ComfyUI + NSFW Checkpoints

• Setup: ComfyUI with models from Civitai
• Popular checkpoints:
  • Juggernaut XL — photorealistic, versatile
  • Pony Diffusion V6 XL — anime/stylized
  • RealVisXL — photorealism focused
• Requirements: 8GB+ VRAM (SDXL-based)
• Pros: Huge community, thousands of LoRAs/checkpoints
• Cons: Complex workflow setup, manual model management

Civitai Models

• Source: civitai.com
• Pros: Largest model repository, community ratings, preview images
• Cons: Need to download individually, manage versions, run locally

When to Use Local vs This Skill

This skill is the easiest way to generate NSFW AI images — no GPU setup, no model downloads, no dependency hell. If you already have a powerful GPU and enjoy tinkering, the local options above are great. If you want results in 2 minutes, install this skill.

🛠️ Requirements

• Python 3.10+
• uv — fast Python package runner
• Atlas Cloud API key — get one free

Dependencies (requests, pillow) are managed automatically by uv — no manual pip install needed.

📁 Project Structure

nsfw-ai-image-generator-skill/
├── SKILL.md              # Skill metadata (used by Claude Code)
├── scripts/
│   └── generate.py       # Main generation script
├── .env.example          # Environment variable template
├── LICENSE               # MIT License
├── README.md             # This file
├── README_zh.md          # 中文说明
├── README_ja.md          # 日本語ドキュメント
└── README_ko.md          # 한국어 문서