T2I-Adapter: Conditioning Adapters

ComfyUI Wiki

T2I-Adapter and CoAdapter conditioning models by TencentARC, providing canny, depth, sketch, pose, color, style, and segmentation conditioning for SD1.4/SD1.5 in ComfyUI.

T2I-Adapter and CoAdapter are conditioning adapter models from TencentARC that provide plug-and-play conditioning for Stable Diffusion 1.4 and 1.5. They enable precise control over image generation using various input conditions including canny edges, depth maps, sketches, poses, color palettes, style references, and segmentation maps.

Variants

T2I-Adapter (Original)

  • Canny (SD1.4, SD1.5 v2) — edge detection conditioning
  • Color (SD1.4) — color palette conditioning
  • Depth (SD1.4, SD1.5 v2) — depth map conditioning
  • Keypose (SD1.4) — pose keypoint conditioning
  • OpenPose (SD1.4) — skeleton pose conditioning
  • Segmentation (SD1.4) — segmentation map conditioning
  • Sketch (SD1.4, SD1.5 v2) — sketch conditioning
  • Style (SD1.4) — style reference conditioning
  • ZoeDepth (SD1.5) — ZoeDepth-based depth conditioning

CoAdapter (Updated)

  • Canny, Color, Depth, Sketch, Style (SD1.5)
  • Fuser — multi-conditioning fusion adapter

Usage in ComfyUI

Place the model files in ComfyUI/models/controlnet/ (the split_files/controlnet/ directory maps to this). Compatible with ComfyUI's standard ControlNet loader nodes.

DeveloperTencentARC
Release Date2023-2024 (repackaged 2025)
ArchitectureT2I-Adapter / CoAdapter
LicenseApache-2.0
TypeConditioning Adapters
Compatible ModelsSD1.4, SD1.5
Total Variants18 adapters

Guides and workflows related to this model series.

No articles found.

Comments

Sign in with GitHub to join the discussion.

Loading comments…