Originally posted to Huggingface.
Full Flux Tools Guide, including example workflows, live now!
If you're looking to get started using Flux, check out our Guide here.
FLUX.1 Depth [dev] is a 12 billion parameter rectified flow transformer capable of generating an image based on a text description while following the structure of a given input image. For more information, please read our blog post.
Key Features
Cutting-edge output quality.
It blends impressive prompt adherence with maintaining the structure of source images based on depth maps.
Trained using guidance distillation, making
FLUX.1 Depth [dev]more efficient.Open weights to drive new scientific research, and empower artists to develop innovative workflows.
Generated outputs can be used for personal, scientific, and commercial purposes as described in the
FLUX.1 [dev]Non-Commercial License.
Usage
We provide a reference implementation of FLUX.1 Depth [dev], as well as sampling code, in a dedicated github repository. Developers and creatives looking to build on top of FLUX.1 Depth [dev] are encouraged to use this as a starting point.
Description
FAQ
Comments (13)
But how to use it in comfy?
Read the Flux Tools Guide. It's linked at the top of the page.
Make sure you update to the latest version or you'll get an error.
It's just like an img2img workflow. You use a preprocessor to create the depth map out of the first image and use that as the latent to generate the new image.
download the workflow. any missing nodes you can install with the comfyui manager
How good is this with close up face portraits?
I uploaded an example for you - https://civitai.com/images/41513359
It depends on the depth preprocessor really, here I'm using MiDAS at 768 resolution. Not bad!
Can someone convert this fp16 model to fp8?
Details
Files
Available On (2 platforms)
Same model published on other platforms. May have additional downloads or version variants.