Anima & Krea-2 INT8 Text-to-Image Workflows Collection
This repository provides optimized ComfyUI workflows for the Anima and Krea-2 model families, utilizing INT8 quantized variants. These workflows enable high-speed, high-fidelity Text-to-Image generation with a low VRAM footprint, while also providing advanced control and image editing capabilities for Anima.
Workflow Files & Overview
The repository contains the following production-ready workflow files.
(Note: YYMM.x represents the release year, month, and version number, e.g., 2606.1 or 2607.1.)
| # | File Name | Workflow Target | Key Features & Quick Notes |
|---|
| 1 | Anima_int8-javanoYYMM.x.json | Anima Text-to-Image (INT8) | INT8 image generation with VNCCS, LLLite-based Pose/Canny/Depth control, SAM 3.1 automatic masking, and inpainting. |
| 2 | Krea2_int8_t2i-javanoYYMM.x.json | Krea-2 Text-to-Image (INT8) | Highly optimized INT8 workflow for Krea-2, designed for fast and efficient image generation. |
Detailed Workflow Breakdown
1. Anima Text-to-Image / Control / Inpainting (Anima INT8)
Designed for high-quality static image generation using the INT8 version of Anima, with additional structural control and automated image editing.
- INT8 Inference: Reduced VRAM usage and improved inference efficiency.
- Standard Text-to-Image: Generate images directly from text prompts using Anima.
- VNCCS Control: Provides Pose Doll-based character and pose control.
- LLLite Control: Supports structural guidance using Pose, Canny, and Depth conditioning.
- SAM 3.1 Masking: Automatically generates masks for selected subjects or regions.
- Inpainting: Combines SAM 3.1 automatic masks with Anima LLLite conditioning for targeted image regeneration.
Supported control workflows include:
- Standard Text-to-Image
- Pose Doll
- Canny
- Depth
- Automatic subject masking
- Region-based inpainting
2. Krea-2 Text-to-Image (Krea-2 INT8)
A dedicated INT8 workflow designed to utilize the aesthetic capabilities of Krea-2 within an efficient quantized framework.
- INT8 Acceleration:
Krea2_int8_t2i-javanoYYMM.x.json leverages INT8 quantization for efficient inference.
- Cinematic & Spatial Optimization: Configured to maximize Krea-2's strengths in lighting, realistic rendering, and composition.
- Efficient Inference: Quantization reduces memory requirements while maintaining high image quality.
Anima Control & Editing
VNCCS
Used for Pose Doll-based character and pose control.
Anima LLLite
Provides lightweight structural conditioning for Anima, including:
- Pose
- Canny
- Depth
- Inpainting
SAM 3.1
Used for automatic segmentation and mask generation.
Typical inpainting pipeline:
Input Image → SAM 3.1 → Automatic Mask → Anima LLLite → Anima INT8
This allows specific subjects or regions to be automatically masked and regenerated.
Prerequisites & Installation
To ensure all workflows load correctly without errors, please ensure your environment is fully updated:
-
ComfyUI: Use a recent version of ComfyUI.
-
Custom Nodes:
- Ensure required quantized model loaders and utility nodes such as
ComfyUI-KJNodes are fully updated via ComfyUI Manager.
- Install ComfyUI_VNCCS for Anima Pose Doll control.
- If any supporting node appears red upon loading, use ComfyUI Manager → Install Missing Custom Nodes.
-
Required Models:
- Anima INT8: Download the Anima INT8 model.
- Anima LLLite: Download the required Anima LLLite control models.
- SAM 3.1: Download the SAM 3.1 model for automatic masking.
- Krea-2 INT8: Download the Krea-2 INT8 model required by the workflow.
Required Models & Resources
Anima INT8
Anima LLLite
SAM 3.1
VNCCS
Krea-2 INT8
Use the appropriate Krea-2 INT8 model required by the workflow.
Workflow Summary
Anima INT8
Anima_int8-javanoYYMM.x.json
Text-to-Image + Pose Doll + Canny + Depth + SAM 3.1 Inpainting
Krea-2 INT8
Krea2_int8_t2i-javanoYYMM.x.json
High-speed Krea-2 INT8 Text-to-Image
Notes
- The previous standard Anima workflow has been replaced by the Anima INT8 workflow.
- The previous Krea-2 GGUF workflow has been removed.
- Anima INT8 supports standard image generation as well as VNCCS, LLLite, and SAM 3.1-based workflows.
- VNCCS is used for Pose Doll control.
- LLLite provides Pose, Canny, and Depth conditioning.
- SAM 3.1 provides automatic mask generation for image editing and inpainting.
- Performance depends on GPU architecture, VRAM, resolution, sampling settings, and model quantization.
License
Please follow the respective licenses of Anima, Krea-2, the quantized model variants, ComfyUI, VNCCS, Anima LLLite, SAM 3.1, and any other third-party models or custom nodes used by these workflows.
