Every model, one typed interface

110+ curated nodes: image, video, language and audio models, plus media utilities. Each one carries typed ports, real defaults and a pinned version.

Read the run-a-node guide

The node catalog

All 117Generate Video 18Generate Images 34Edit & Analyze Image 32Upscale Image 5Generate Audio 8Transcribe Audio 2Language 113D 1Media utilities 6

Edit & Analyze Image

Edit, cut out, segment or measure a picture you already have.

Flux Kontext Pro input
Flux Kontext Pro output

Flux Kontext Pro

Instruction edit holding identity and style, at Kontext's pro quality tier

Qwen Image Edit Plus input
Qwen Image Edit Plus output

Qwen Image Edit Plus

Instruction editing with strong text rendering inside the image.

Flux Fill Pro input
Flux Fill Pro output

Flux Fill Pro

Inpaint a masked region from a prompt at pro quality

Background Remover input
Background Remover output

Background Remover

Remove the image background → transparent PNG.

BiRefNet (Cutout) input
BiRefNet (Cutout) output

BiRefNet (Cutout)

High-resolution background removal with fine edge matting → transparent PNG.

Segment Anything 2 (Masks) input
Segment Anything 2 (Masks) output

Segment Anything 2 (Masks)

Segment every region of an image into one combined mask.

Florence-2 (Caption & Detect) input

{'<DETAILED_CAPTION>': 'The image shows a golden retriever puppy sitting in a field of daisies, surrounded by lush green grass and a bright blue sky.'}

Florence-2 (Caption & Detect)

Caption, tag or read text out of an image.

Depth Anything 3 Base input
Depth Anything 3 Base output

Depth Anything 3 Base

Estimate depth from a single image → depth map (near = bright) with a confidence map and estimated camera intrinsics.

Bria Eraser (Remove Object) input
Bria Eraser (Remove Object) output

Bria Eraser (Remove Object)

Erase the masked region and fill it from its surroundings.

Bria Expand (Outpaint) input
Bria Expand (Outpaint) output

Bria Expand (Outpaint)

Extend an image past its edges to fill a wider canvas.

Color Matcher input
Color Matcher output

Color Matcher

Match one image's palette to a reference, or fix white balance.

ControlNet Pose input
ControlNet Pose output

ControlNet Pose

Generate an image from a prompt while preserving a detected human pose.

ControlNet Scribble input
ControlNet Scribble output

ControlNet Scribble

Generate a detailed image from a prompt while preserving a scribbled composition.

Depth Anything 3 Mono input
Depth Anything 3 Mono output

Depth Anything 3 Mono

Best-quality single-image depth → depth map (near = bright) plus a sky mask.

Depth Anything 3 Small input
Depth Anything 3 Small output

Depth Anything 3 Small

Estimate depth from a single image → depth map (near = bright) with a confidence map and estimated camera intrinsics.

Depth Anything V1 Base input
Depth Anything V1 Base output

Depth Anything V1 Base

Estimate depth from a single image → a depth map (near = bright) plus a colored preview.

Depth Anything V1 Large input
Depth Anything V1 Large output

Depth Anything V1 Large

Estimate depth from a single image → a depth map (near = bright) plus a colored preview.

Depth Anything V1 Small input
Depth Anything V1 Small output

Depth Anything V1 Small

Estimate depth from a single image → a depth map (near = bright) plus a colored preview.

Depth Anything v2 (Depth Map) input
Depth Anything v2 (Depth Map) output

Depth Anything v2 (Depth Map)

Read a depth map off any image, ready for depth-guided generation.

Depth Anything V2 Small input
Depth Anything V2 Small output

Depth Anything V2 Small

Estimate depth from a single image → a depth map (near = bright) plus a colored preview.

FireRed Image Edit 1.1 input
FireRed Image Edit 1.1 output

FireRed Image Edit 1.1

High-fidelity instruction editing from one or two references

Flux Canny input
Flux Canny output

Flux Canny

Regenerate an image while preserving exact outlines and silhouettes from the reference.

Flux Depth input
Flux Depth output

Flux Depth

Regenerate an image while preserving depth/perspective structure from the reference.

Flux Fill (Inpaint) input
Flux Fill (Inpaint) output

Flux Fill (Inpaint)

Fill masked areas of an image guided by a text prompt. White mask = fill, black = keep.

Flux Kontext input
Flux Kontext output

Flux Kontext

Edit an image with text instructions while preserving character identity and style consistency.

Flux Redux (Variations) input
Flux Redux (Variations) output

Flux Redux (Variations)

Generate style-consistent variations of an image. No text prompt required.

Grounded SAM (Text to Mask) input
Grounded SAM (Text to Mask) output

Grounded SAM (Text to Mask)

Turn a plain-text description of a subject into a cutout mask.

Grounding DINO (Detect) input
Grounding DINO (Detect) output

Grounding DINO (Detect)

Locate objects named in plain text, as boxes plus an annotated image.

IC-Light (Relight) input
IC-Light (Relight) output

IC-Light (Relight)

Relight a subject to match a described lighting setup.

Pruna Image Edit (Fast) input
Pruna Image Edit (Fast) output

Pruna Image Edit (Fast)

Sub-second instruction editing, with task presets and a second reference image

Pruna Try-On input
Pruna Try-On output

Pruna Try-On

Dress a person photo in up to three garments, holding identity

Segment Anything 2 (Video)

Track a clicked subject through a clip and mask it every frame

Run one from your own code

Open the studio