<h2 id="comfyui-anima-suite">ComfyUI Anima Suite</h2>A dedicated custom node suite and ready-to-use workflow designed for Anima 3.8B generation in ComfyUI. It packages two-stage sampling, automatic LoRA block remapping, aspect ratio presets, and built-in offline Danbooru tag, character, and artist browsers into a compact canvas setup.Everything runs 100% locally with zero external network requests.Download size is ~4.6 GB because it bundles a complete offline library of 78,000+ local artist and character reference thumbnails.<hr /><h2 id="quick-setup">Quick Setup</h2><ol><li>Copy or extract this folder into <code>ComfyUI/custom_nodes/comfyui-anima-suite/</code>.</li><li>Restart ComfyUI.</li><li>Load the bundled workflow: <code>example_workflows/Anima_delta_mix.json</code>.</li><li>Enter your prompt (or pick from the tag/artist drawers) and queue!</li></ol>Required Model Files:<ul><li>Checkpoint: Anima model (<code>models/checkpoints/</code>)</li><li>Text Encoder: Compatible text encoder (<code>models/text_encoders/</code>)</li><li>VAE: Compatible image VAE (<code>models/vae/</code>)</li><li>LoRAs (Optional): Any Anima-compatible LoRA (<code>models/loras/</code>)</li></ul><hr /><h2 id="usage-and-recommended-settings">Usage &amp; Recommended Settings</h2>Sampler: euler_ancestral Scheduler: simpleDual-Pass Setup<ul><li>Pass 1 (Base): Steps: <code>10 – 15</code> | CFG: <code>4.0 – 5.0</code> | Denoise: <code>1.00</code></li><li>Pass 2 (Hires Refine): Steps: <code>10 – 15</code> | CFG: <code>2.0 – 4.0</code> | Denoise: <code>0.20 – 0.60</code> Tip: Lower denoise (<code>0.20 – 0.30</code>) keeps composition identical to Pass 1; higher denoise (<code>0.45 – 0.60</code>) adds micro-details and change some details.</li></ul>Single-Pass Setup<ul><li>Steps: <code>25 – 35</code> | CFG: <code>3.0 – 5.0</code> | Denoise: <code>1.00</code> (disable Pass 2 via the <code>run_pass_2</code> toggle on the sampler)</li></ul><hr />Key Features<ul><li>Anima Dual Sampler: Unified two-stage KSampler with shared seed coordination, pass toggles (<code>run_pass_1</code> / <code>run_pass_2</code>), and an option to refine with or without LoRAs applied.</li><li>Anima LoRA Stack: 4-slot stack featuring automatic block remapping across 28, 40, and 52-block Anima architectures to match your active checkpoint.</li><li>Anima Prompt Hub &amp; Resolution: Single-pipe pass-through wiring, collapsible negative drawer, and aspect ratio presets (1:1, 4:5, 3:4, 2:3, 9:16, etc.) with live dimension calculation.</li><li>Built-in Tag &amp; Artist Drawers: Searchable offline browser with ~36,000 characters and ~42,000 artists (with thumbnails), plus a Danbooru tag helper.</li><li>Prompt History &amp; Image Inspector: Drop any generated image onto the canvas to read embedded metadata back into nodes, with automatic prompt version saving.</li></ul><hr />Tested system:GPU: NVIDIA GeForce RTX 5070 (12 GB VRAM) --- RAM: 32 GBGeneration Speed (2:3 aspect ratio / 832×1216 base):Dual-Pass Full High-Res (1.5× upscale): ~35 seconds totalPass 1 (11 steps): ~8 seconds --- Pass 2 (15 steps): ~25 seconds<hr /> Drop a review or showcase your generations below, and let me know in the comments if you run into any issues or want new features added!

Files with SHA256: dc249fc8ee8db1852757e46b5e4c2331d784fa4c16d52ce4f778efc2fdd14d8b