SDXL
Stability AI's flagship text-to-image model, delivering dramatically improved image quality, composition, and detail over previous Stable Diffusion versions.
Stability 
WaveSpeed Text-to-image generation with detailed composition control
SDXL (Stable Diffusion XL) is an open-source image generation model developed by Stability AI and released in July 2023. It accepts text prompts and optional image inputs to produce images, and supports workflows including text-to-image generation, image-to-image editing, and inpainting. The model is available in two configurations: SDXL 1.0, optimized for out-of-the-box use, and SDXL 1.0 Open, which allows fine-tuning with custom data and inference code. Both variants are deployable via AWS SageMaker and Amazon Bedrock.
SDXL is designed for designers, creative professionals, and developers who need generative imagery at scale. Its open model variant supports customization through fine-tuning, making it usable for specialized image pipelines beyond general-purpose prompting. Inputs include image URLs, numeric parameters for dimensions, and a seed value for reproducible outputs. The model is tagged as open source and multi-modal, reflecting its support for both text and image inputs.
What SDXL supports
Text-to-Image Generation
Generates images from natural language text prompts, supporting photorealistic and artistic styles with control over composition and detail.
Image-to-Image Editing
Transforms or refines an existing image using a source image URL and text guidance as inputs.
Inpainting
Fills or replaces specific regions within an image based on text prompts, enabling targeted edits to existing visuals.
Seed-Based Reproducibility
Accepts a seed value as input so that identical prompts and settings produce consistent, reproducible image outputs.
Open Fine-Tuning
The SDXL 1.0 Open variant supports fine-tuning with custom datasets and custom inference code for specialized use cases.
Dimension Control
Accepts numeric width and height parameters as inputs, allowing explicit control over output image dimensions.
Ready to build with SDXL?
Get Started FreeCommon questions about SDXL
What is the context window for SDXL?
The metadata lists a context window of 10,000 tokens for SDXL, which governs the length of text prompt input the model can process.
Is SDXL open source?
Yes. SDXL is tagged as open source. The SDXL 1.0 Open variant specifically allows developers to fine-tune the model with custom data and run custom inference code.
What input types does SDXL accept?
SDXL accepts an image URL (for image-to-image or inpainting workflows), two numeric values (typically width and height), and a seed value for reproducible generation.
Where can SDXL be deployed?
According to the metadata, SDXL is available for deployment via AWS SageMaker and Amazon Bedrock, in addition to Stability AI's own platform API.
When was SDXL trained and released?
The model's training date is listed as July 2023, which aligns with Stability AI's public announcement of SDXL 1.0.
What are the two deployment configurations for SDXL?
SDXL is available as SDXL 1.0, optimized for speed and quality out of the box, and SDXL 1.0 Open, which provides full access for fine-tuning and extending the model for specialized use cases.
What people think about SDXL
Community discussion around SDXL on Reddit reflects active use among developers and hobbyists, with the fine-tuning process being a frequently discussed topic. A highly upvoted thread detailing the costs and challenges of fine-tuning SDXL drew significant engagement, indicating that customization workflows are a common area of interest.
Users also raise concerns about the resource costs and technical complexity involved in fine-tuning SDXL, with one thread documenting $16,000 in wasted compute as a cautionary account. Broader community threads show ongoing interest in comparing SDXL to newer text-to-image models, suggesting users actively evaluate it alongside other options in the ecosystem.
AMD and Stability AI release Stable Diffusion for AMD NPUs
The Gory Details of Finetuning SDXL and Wasting $16k
Just another Wan 2.1 14B text-to-image post
Wan 2.2 - We've barely showcased its potential
Parameters & options
A specific value that is used to guide the 'randomness' of the generation.
Explore similar models
Stability SDXL LoRA
Stability AI's SDXL LoRA is a powerful text-to-image model combining a 3.5 billion parameter architecture with flexible LoRA customization for professional-grade, high-resolution image generation.
Stability Stable Diffusion 3
Stable Diffusion 3 is Stability AI's most advanced text-to-image model, featuring a new Multimodal Diffusion Transformer architecture that delivers superior prompt adherence, typography, and visual quality.
Stability Stable Image Ultra
Stability AI's most advanced text-to-image model, delivering exceptional photorealism and creative detail from short text prompts.
Ideogram Ideogram Upscale
Ideogram's AI-powered image upscaler that enhances photos and generated images to stunning 8K resolution with sharp, production-ready detail.
X.ai Grok Imagine Pro
xAI's premium image generation model, delivering higher-quality visuals than the standard grok-imagine-image.
Black Forest Labs FLUX.1 [dev] LoRA
FLUX.1 [dev] with LoRA support for fast, high-quality, personalized image generation using pre-trained style adapters.
Start building with SDXL
No API keys required. Create AI-powered workflows with SDXL in minutes — free.