DeepFloyd

by DeepFloyd and StabilityAI and LAION (major contributors)

About the model:

Pixel-based triple-cascaded diffusion model.

Model type:

Image

Model performance class:

Full

Link to the model:

https://huggingface.co/DeepFloyd/IF-I-XL-v1.0

Base models:

(undefined)

End model:

IF-I-XL-V1.0

End model license:

DeepFloyd IF Research License Agreement

About the organisation:

Collaboration between various organizations.

Link to the organisation:

[ "https://huggingface.co/DeepFloyd", "https://stability.ai/", "https://laion.ai/" ]

Model release date:

April 2023

Availability

Base Model Data

Model training partially based on LAION-A, however also uses internal datasets.

https://huggingface.co/DeepFloyd/IF-I-XL-v1.0#training

End User Model Data

Base Model Weights

Gated model made available through HuggingFace.

https://huggingface.co/DeepFloyd/IF-I-XL-v1.0

End User Model Weights

Training Code

No training code found.

https://github.com/deep-floyd/IF

Documentation

Code Documentation

No training code, so no documentation.

Hardware Architecture

Hardware information provided.

https://huggingface.co/DeepFloyd/IF-I-XL-v1.0

Preprint

Model inspired by paper, however does not have a paper of its own.

https://arxiv.org/pdf/2205.11487

Paper

No peer-reviewed paper found.

Modelcard

Model card provides the desired information.

https://huggingface.co/DeepFloyd/IF-I-XL-v1.0

Datasheet

No datasheet found.

Access

Licenses

DeepFloyd IF Research License Agreement, not an OSI-approved license.

https://huggingface.co/DeepFloyd/IF-I-XL-v1.0#model-details

Is this information not up to date?

Contribute here ->

Supported by the Centre for Language Studies and the Dutch Research Council. Website design & development © 2024 by BSTN. This version of the index generated 09 April 2026, website content last updated 11 March 2026.