<div data-youtube-video="https://youtu.be/EqhYXn8oDqk"><iframe src="https://www.youtube.com/embed/EqhYXn8oDqk" width="853" height="480" allowfullscreen></iframe></div><br /><br />
<p>RunningHub workflow page: https://www.runninghub.ai/zh-cn/post/2089668746540572673?inviteCode=rh-v1111</p>
<h2>Anima 2.9B Three-Stage High-Quality Render</h2>
<p>Purpose: a three-pass Anima 2.9B render. The base image is generated, then two latent upscales with low-denoise passes progressively resolve detail, so the final image is built through three sampling stages rather than one.</p>
<h2>Graph (as saved)</h2>
<p>16 nodes / 23 links.</p>
<ul>
<li>UNETLoader <code>Anima-2.9B-preview-v1.safetensors</code>; CLIPLoader <code>qwen_3_06b_base.safetensors</code> (type <code>stable_diffusion</code>); VAELoaderKJ <code>qwen_image_vae.safetensors</code>.</li>
<li>Base canvas: EmptyLatentImage 1024x1024 + FluxResolutionNode at 1.0 MP.</li>
<li>Three KSamplers, all 28 steps / CFG 5 / <code>res_multistep</code> / <code>linear_quadratic</code>: stage 1 denoise 1, stage 2 denoise 0.35, stage 3 denoise 0.3.</li>
<li>Between stages: two <code>LatentUpscaleBy</code> nodes, nearest-exact at 1.25x each.</li>
<li>SaveImage prefix <code>Anima_2.9B/three_stage_hq</code>; a PreviewImage is kept in the graph for the intermediate stage.</li>
</ul>
<h2>How to use</h2>
<p>The two 0.35 / 0.3 denoise values are the main tuning surface: raise them if the upscaled stages look soft, lower them if the model starts repainting the subject. Changing the two 1.25x factors changes the final size, so re-check the output resolution after editing them.</p>
<h2>Limits and not-verified items</h2>
<p>No inference was run for this listing: no speed, VRAM or quality claim is made; values are quoted from the saved nodes. Weights, VAE, encoder and demo images are not bundled. The cover image is a promotional still, not a frame produced by this graph.</p>
<p>Tutorial: <a href="https://www.youtube.com/watch?v=EqhYXn8oDqk" rel="ugc">YouTube</a> · <a href="https://www.bilibili.com/video/BV1wG8G62Epx" rel="ugc">Bilibili</a>. Workflows supplied through AI-KSK; verify workflow, node, model, encoder, image and cover rights separately. No blanket commercial or redistribution license is asserted.</p>

Anima 2.9B Three-Stage High-Quality Render

<div data-youtube-video="https://youtu.be/EqhYXn8oDqk"></div><br /><br />
<p>RunningHub workflow page: https://www.runninghub.ai/zh-cn/post/2089668746540572673?inviteCode=rh-v1111</p>
Anima 2.9B Three-Stage High-Quality Render
<p>Purpose: a three-pass Anima 2.9B render. The base image is generated, then two latent upscales with low-denoise passes progressively resolve detail, so the final image is built through three sampling stages rather than one.</p>
Graph (as saved)
<p>16 nodes / 23 links.</p>
<ul>
<li>UNETLoader <code>Anima-2.9B-preview-v1.safetensors</code>; CLIPLoader <code>qwen_3_06b_base.safetensors</code> (type <code>stable_diffusion</code>); VAELoaderKJ <code>qwen_image_vae.safetensors</code>.</li>
<li>Base canvas: EmptyLatentImage 1024x1024 + FluxResolutionNode at 1.0 MP.</li>
<li>Three KSamplers, all 28 steps / CFG 5 / <code>res_multistep</code> / <code>linear_quadratic</code>: stage 1 denoise 1, stage 2 denoise 0.35, stage 3 denoise 0.3.</li>
<li>Between stages: two <code>LatentUpscaleBy</code> nodes, nearest-exact at 1.25x each.</li>
<li>SaveImage prefix <code>Anima_2.9B/three_stage_hq</code>; a PreviewImage is kept in the graph for the intermediate stage.</li>
</ul>
How to use
<p>The two 0.35 / 0.3 denoise values are the main tuning surface: raise them if the upscaled stages look soft, lower them if the model starts repainting the subject. Changing the two 1.25x factors changes the final size, so re-check the output resolution after editing them.</p>
Limits and not-verified items
<p>No inference was run for this listing: no speed, VRAM or quality claim is made; values are quoted from the saved nodes. Weights, VAE, encoder and demo images are not bundled. The cover image is a promotional still, not a frame produced by this graph.</p>
<p>Tutorial: <a href="https://www.youtube.com/watch?v=EqhYXn8oDqk" rel="ugc">YouTube</a> · <a href="https://www.bilibili.com/video/BV1wG8G62Epx" rel="ugc">Bilibili</a>. Workflows supplied through AI-KSK; verify workflow, node, model, encoder, image and cover rights separately. No blanket commercial or redistribution license is asserted.</p>

03_Anima-2.9B________.json

v1.0

<div data-youtube-video="https://youtu.be/ir3_1v9LMTA"><iframe src="https://www.youtube.com/embed/ir3_1v9LMTA" width="853" height="480" allowfullscreen></iframe></div><br />
<br />
AI-KSK ComfyUI workflow for three-stage high-quality rendering with the Anima 2.9B model.<br />
<br />
Main features:<br />
- Anima 2.9B workflow with the connected nodes preserved from the source JSON.<br />
- Three-stage high-quality rendering.<br />
- Designed for reproducible ComfyUI and RunningHub use.<br />
<br />
Suggested workflow:<br />
1. Download the workflow JSON and import it into ComfyUI.<br />
2. Install the required model files and custom nodes shown in the workflow.<br />
3. Adjust the prompt, resolution, and output settings, then run the connected output path.<br />
<br />
RunningHub workflow:<br />
https://www.runninghub.ai/post/2089343588722647041?inviteCode=rh-v1111<br />
<br />
Bilibili tutorial:<br />
https://www.bilibili.com/video/BV1wG8G62Epx/<br />
<br />
AI-KSK resources and tutorials:<br />
https://space.bilibili.com/110353151<br />
https://ko-fi.com/aiksk

Anima 2.9B Three-Stage Render

<div data-youtube-video="https://youtu.be/ir3_1v9LMTA"></div><br />
<br />
AI-KSK ComfyUI workflow for three-stage high-quality rendering with the Anima 2.9B model.<br />
<br />
Main features:<br />
- Anima 2.9B workflow with the connected nodes preserved from the source JSON.<br />
- Three-stage high-quality rendering.<br />
- Designed for reproducible ComfyUI and RunningHub use.<br />
<br />
Suggested workflow:<br />
1. Download the workflow JSON and import it into ComfyUI.<br />
2. Install the required model files and custom nodes shown in the workflow.<br />
3. Adjust the prompt, resolution, and output settings, then run the connected output path.<br />
<br />
RunningHub workflow:<br />
https://www.runninghub.ai/post/2089343588722647041?inviteCode=rh-v1111<br />
<br />
Bilibili tutorial:<br />
https://www.bilibili.com/video/BV1wG8G62Epx/<br />
<br />
AI-KSK resources and tutorials:<br />
https://space.bilibili.com/110353151<br />
https://ko-fi.com/aiksk

Files with SHA256: bd029074420e71e12f2403bc686c6453912674536d762154de82583ab6abf00b