SeedVarianceEnhancer
A ComfyUI custom node that adds diversity to Z-Image Turbo outputs by adding random noise to text embeddings.
5,383,139 AI resources indexedWorking toward the world’s largest AI resource index.
AI software for creative and professional work.
Local Inference Engines RAG and Retrieval AI Evaluation Tools AI Monitoring
A ComfyUI custom node that adds diversity to Z-Image Turbo outputs by adding random noise to text embeddings.
Vision-aware text conditioning for the Krea2 / K2 model. Feeds reference images through the Qwen3-VL-4B vision path with the Krea2 descriptor template; optional per-image mask crops to the masked region. No VAE (Krea2 has no reference-latent pathway). Auto-growing image+mask slots.
ComfyUI native nodes for IC-Light with a Video node
ComfyUI port of Vectorscope CC and Diffusion Color Grading by Haoming02. Makes it possible to adjust Brightness/Contrast/Saturation/Hue during image generation.
Manipulate the details of generations
Color Grading for Stable Diffusion
H100-optimized face detection with Auto-Padding, Temporal Smoothing, Aspect Ratio Presets, and full batch processing for LTX-Video avatar pipelines. Backward compatible with v1.x workflows.
A ComfyUI node that takes multiple images, scales them, VAE encodes them, and chains them as reference latents into a single conditioning output. Useful for edit models like Flux-2-klein that accept multiple reference images.
All-in-One Face Fix KSampler for ComfyUI with YOLO detection and SAM segmentation
Training-free TRELLIS.2 acceleration using HiCache Hermite / HiCache++ DMD (hicache-pp). Insert between the TRELLIS2 loader and sampler. Earlier RTX 5090 README result (1.9x) is a setup-specific historical measurement, not a current-release guarantee; see README for metrics and limits.
Multi-reference image and optional AVref audio conditioning nodes for LTX-2.5 in ComfyUI.
ComfyUI loaders for MiniMax-H3 NF4 weights, plus prompt, canvas, audio reroll, and AV latent tools.
Try a broader search or clear some filters.
Platform labels describe intended use. Read each resource's author credits, requirements and available access methods.