Sold by: Stability AI
Deployed on AWS
Get a preview of SDXL from Stability AI, the leading model for creating and editing images.
Overview
SDXL Beta 0.9 offers a preview to the largest open source image model from Stability AI. SDXL produces high resolution images at native 1024px resolution. With SDXL Beta 0.9, Stability introduces a brand new 2-stage architecture for image generative AI models, and integrates the largest CLIP model. The result: stunning quality in image generation with minimal need for prompt engineering.
Highlights
- The foundation model for images: All text to image, image to image, inpainting, and outpainting workflows are handled by the official SDXL model. SDXL introduces a new SOTA architecture for image generation, comprising a 3.5B parameter base model stage and a 6.6B parameter ensemble pipeline.
- Unprecedented quality and ease of use: Native 1024x1024 image generation with cinematic photorealism and fine detail. The most advanced text generation within images. Fine-tuned to create complex compositions with basic natural language prompting, thanks to the largest CLIP model in production.
- Beta: SDXL 0.9 is a beta version of SDXL 1.0. While it is a powerful image generation model, for increased professionalism in image output, consider using SDXL 1.0
Details
Sold by
Categories
Delivery method
Latest version
Deployed on AWS
New
Introducing multi-product solutions
You can now purchase comprehensive solutions tailored to use cases and industries.
Features and programs
Trust Center
Access real-time vendor security and compliance information through their Trust Center powered by Drata or Vanta. Review certifications and security standards before purchase.
Financing for AWS Marketplace purchases
AWS Marketplace now accepts line of credit payments through the PNC Vendor Finance program. This program is available to select AWS customers in the US, excluding NV, NC, ND, TN, & VT.
Pricing
This product is available free of charge. Free subscriptions have no end date and may be canceled any time.
Additional AWS infrastructure costs may apply. Use the AWS Pricing Calculator to estimate your infrastructure costs.
Dimensions summary
This listing is free to use, so the dimensions reflect how you run the model rather than license fees. Four dimensions charge by host hour and differ by the compute instance you pick. One instance runs batch inference, where you process requests in groups. Three run real-time inference, which responds to requests as they arrive. Each instance type offers different compute and GPU capacity, so your choice scales cost with the hardware you select. A fifth dimension bills per inference request instead of by host hour, giving you a usage-based option tied to request volume.
Top-of-mind questions for buyers
A host hour is one hour of running time on the compute instance you deploy. Each instance type (ml.g4dn.12xlarge, ml.p4de.24xlarge, ml.p4d.24xlarge, ml.g5.2xlarge) carries its own hourly rate. Charges accrue while the instance runs, not by how many images you generate.
Host-hour dimensions meter running time on a chosen instance, so cost follows how long the hardware stays active. The per-request dimension (inference.count.m.i.c) meters each inference call instead. Request-based billing tracks volume of calls; host-hour billing tracks active instance time regardless of request count.
The software itself is listed as free, so there are no license fees. You still pay the underlying compute cost per the host-hour or per-request dimensions shown. The vendor's own terms note self-hosting is free for organizations under $1M in annual revenue; larger commercial users should contact the vendor.
stability.ai
Helpful?
Vendor refund policy
n/a
Legal
Vendor terms and conditions
Upon subscribing to this product, you must acknowledge and agree to the terms and conditions outlined in the vendor's End User License Agreement (EULA) .
Content disclaimer
Vendors are responsible for their product descriptions and other product content. AWS does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.
Delivery details
Amazon SageMaker model
An Amazon SageMaker model package is a pre-trained machine learning model ready to use without additional training. Use the model package to create a model on Amazon SageMaker for real-time inference or batch processing. Amazon SageMaker is a fully managed platform for building, training, and deploying machine learning models at scale.
Deploy the model on Amazon SageMaker AI using the following options:
Version release notes
This release resolves a bug with mask image inputs.
Additional details
Inputs
- Summary
This model accepts JSON input aligned with the Stability REST API as well as protocol buffers aligned with the Stability GRPC API .
- Limitations for input type
- Input images must be base64 encoded in JSON or supplied via protobuf.
- Input MIME type
- application/json, application/x-protobuf
{
"cfg_scale": 7,
"height": 1024,
"width": 1024,
"steps": 50,
"seed": 42,
"sampler": "K_DPMPP_2M",
"text_prompts": [
{
"text": "A photograph of fresh pizza with basil and tomatoes, from a traditional oven",
"weight": 1
}]
}
https://platform.stability.ai
Input data descriptions
The following table describes supported input data fields for real-time inference and batch transform.
Field name | Description | Constraints | Required |
|---|---|---|---|
height | The height of the image in pixels. Must be a multiple of 64.
On p4d/p4de instances, resolutions values are unlocked and only limited by GPU memory.
On g5 instances, optimized native resolutions are limited to:
1024x1024
1152x896 / 896x1152
1216x832 / 832x1216
1344x768 / 768x1344
1536x640 / 640x1536
Lower resolutions than native will be cropped. Higher resolutions will result in an error. | Default value: 1024
Type: Integer | No |
width | The width of the image in pixels. Must be a multiple of 64.
On p4d/p4de instances, resolutions values are unlocked and only limited by GPU memory.
On g5 instances, optimized native resolutions are limited to:
1024x1024
1152x896 / 896x1152
1216x832 / 832x1216
1344x768 / 768x1344
1536x640 / 640x1536
Lower resolutions than native will be cropped. Higher resolutions will result in an error. | Default value: 1024
Type: Integer | No |
text_prompts | An array of text prompts to use for generation.
Given a text prompt with the text A lighthouse on a cliff and a weight of 0.5, it would be represented as:
"text_prompts": [
{
"text": "A lighthouse on a cliff",
"weight": 0.5
}
] | Type: FreeText
Limitations: Structured JSON array of prompts. | Yes |
cfg_scale | How strictly the diffusion process adheres to the prompt text (higher values keep your image closer to your prompt) | Default value: 7
Type: Continuous
Minimum: 0
Maximum: 35 | No |
sampler | Which sampler to use for the diffusion process. | Default value: auto
Type: Categorical
Allowed values: DDIM,DDPM,K_DPMPP_SDE,K_DPMPP_2M,K_DPMPP_2S_ANCESTRAL,K_DPM_2,K_DPM_2_ANCESTRAL,K_EULER,K_EULER_ANCESTRAL,K_HEUN,K_LMS | No |
seed | Random noise seed (omit this option or use 0 for a random seed) | Default value: 0
Type: Integer | No |
style_preset | Pass in a style preset to guide the image model towards a particular style. This list of style presets is subject to change. | Default value: none
Type: Categorical
Allowed values: enhance,anime,photographic,digital-art,comic-book,fantasy-art,line-art,analog-film,neon-punk,isometric,low-poly,origami,modeling-compound,cinematic,3d-model,pixel-art,tile-texture | No |
steps | Number of diffusion steps to run. | Default value: 30
Type: Integer
Minimum: 10
Maximum: 150 | No |
Resources
Vendor resources
Support
AWS infrastructure support
AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.
Similar products
The early beta for Stability AI’s foundation model for image generation, targeting medium resolution (512px) for fast image generation.
The official foundation model for image generation from Stability. Deploy this optimized instance and serve generative AI within minutes.
One of the most customizable and accessible image models available, maintaining top-tier performance in prompt adherence and image quality.
The StableLogic SFTP Transfer delivers business-grade SFTP security and control directly within your own AWS account, with no legacy managed file transfer platform required. Deploy in minutes via AWS Marketplace, connect any standard SFTP client, and land files directly in your own Amazon S3 bucket with SSE-KMS encryption, per-user access controls, IP whitelisting, and bandwidth throttling. ISO 27001 certified, GDPR compliant, and built for regulated industries, external partner file exchange, and cloud-native data platforms.
Customer reviews
No customer reviews yet
Be the first to review this product . We've partnered with PeerSpot to gather customer feedback. You can share your experience by writing or recording a review, or scheduling a call with a PeerSpot analyst.