NSFW-gen-v2 - AI Language Models Tool

Overview

NSFW-gen-v2 is a text-to-image model published on Hugging Face by UnfilteredAI that is explicitly designed to produce uncensored, adult (NSFW) imagery from text prompts. The model is provided as a Diffusers/StableDiffusionXLPipeline artifact and is distributed in safetensors format; the model card states it was finetuned to produce uncensored outputs and includes prompt-friendly behavior for more realistic "3D" rendering styles. ([huggingface.co](https://huggingface.co/UnfilteredAI/NSFW-gen-v2?not-for-all-audiences=true)) Targeted at creators, researchers, and moderation testers who need unrestricted generation, NSFW-gen-v2 is a mid‑sized image-generation finetune (around 3.47B parameters) optimized to run in FP16 for improved inference efficiency. The Hugging Face page reports ~8.6K downloads in the last 30 days and several community Spaces that use the model, indicating active community adoption — but it is explicitly marked "Not‑For‑All‑Audiences" and carries a nonstandard "other" license so integrators should review platform and legal constraints before use. ([huggingface.co](https://huggingface.co/UnfilteredAI/NSFW-gen-v2?not-for-all-audiences=true))

Model Statistics

• Downloads: 8,672
• Likes: 520
• Pipeline: text-to-image
• Parameters: 3.5B

License: other

Model Details

Architecture & base: NSFW-gen-v2 is published as a StableDiffusionXLPipeline artifact (text-to-image) and the model card identifies a PixelGen-style base/ancestor (listed as HelpingAI/PixelGen or OEvortex/PixelGen in third‑party summaries). The model files are distributed as safetensors and the card lists FP16 as the operational tensor type for inference. Parameter count is reported at ~3.47 billion (3B+ class). ([huggingface.co](https://huggingface.co/UnfilteredAI/NSFW-gen-v2?not-for-all-audiences=true)) Capabilities & controls: The model exposes standard diffusion generation controls (guidance_scale, negative_prompt, num_inference_steps, width, height, seed and compatible schedulers) when used via Diffusers or hosted inference endpoints. The model card highlights a prompt‑triggered "3d" or "3d style" rendering capability to bias outputs toward more realistic 3D-like renderings. The repo listing also indicates language-tagging examples (English, Portuguese, Thai) and that the model is not currently deployed by any official inference provider on the Hugging Face listing. ([ai.azure.com](https://ai.azure.com/catalog/models/unfilteredai-nsfw-gen-v2?utm_source=openai)) Files & deployment: Files on the Hub are safetensors and intended for Diffusers-compatible loading. Because the license is "other" and the model is labeled sensitive, usage may require additional review or access controls depending on the target platform and local laws. The Hugging Face page shows numerous community Spaces and forks that run or demonstrate the model, but the hosting status varies. ([huggingface.co](https://huggingface.co/UnfilteredAI/NSFW-gen-v2?not-for-all-audiences=true))

Key Features

• Uncensored, explicit-capable text-to-image generation using natural language prompts.
• StableDiffusionXL pipeline compatibility for high-fidelity image synthesis.
• FP16-optimized safetensors files to reduce memory footprint during inference.
• ~3.47B-parameter finetune for detailed, diverse image outputs.
• Prompt-driven 3D-style rendering (include "3d" or "3d style" in prompts).
• Multilingual prompt support noted for English, Portuguese, and Thai.

Example Usage

Example (python):

from diffusers import StableDiffusionXLPipeline
import torch

# Note: this model produces explicit NSFW outputs. Use only where lawful and permitted.
model_id = "UnfilteredAI/NSFW-gen-v2"

# Load (may require appropriate access or a Hugging Face token for restricted repos)
pipe = StableDiffusionXLPipeline.from_pretrained(model_id, torch_dtype=torch.float16)
pipe = pipe.to("cuda")

prompt = "A highly detailed, cinematic 3d portrait of an adult fantasy character, dramatic lighting"
result = pipe(
    prompt=prompt,
    negative_prompt="lowres, deformed, watermark, text",
    guidance_scale=7.5,
    num_inference_steps=30,
    width=1024,
    height=1024,
    generator=torch.Generator(device="cuda").manual_seed(42)
)

image = result.images[0]
image.save("nsfw_output.png")

# Adjust guidance_scale, steps, negative_prompt and seed per your quality/speed tradeoffs.
# Respect local law and platform terms when generating explicit content.

Benchmarks

Downloads (last 30 days): 8,672 (Source: https://huggingface.co/UnfilteredAI/NSFW-gen-v2)

Likes / Followers (Hugging Face): ≈520 likes (Source: https://huggingface.co/UnfilteredAI/NSFW-gen-v2)

Model size (parameters): ≈3.47B parameters (Source: https://huggingface.co/UnfilteredAI/NSFW-gen-v2)

Pipeline / Task: StableDiffusionXLPipeline (text-to-image) (Source: https://ai.azure.com/catalog/models/unfilteredai-nsfw-gen-v2)

Tensor type: FP16 (optimized for inference) (Source: https://huggingface.co/UnfilteredAI/NSFW-gen-v2)

Key Information

• Category: Language Models
• Type: AI Language Models Tool