<h1 id="anima29b-p1-+-turbo-1.1-merged-model-(unofficial-merge)">Anima29B P1 + Turbo 1.1 Merged Model (Unofficial Merge)</h1><p><strong>Unofficial community derivative</strong>: a full UNET produced by <strong>block-merging</strong> Anima 2.9B Preview 1 with Anima Turbo 1.1. NOT an official release by Anima / CircleStone Labs.</p><p><strong>Use directly</strong>: load it as the diffusion model — no LoRA or patch nodes needed.</p><ul><li><p>File: <code>anima29b-p1-turbo1.1.safetensors</code></p></li><li><p>Size: ~5.4 GiB — full model (merged UNET, 40 blocks, BF16)</p></li></ul><p>Goal: turbo speed (8 steps) with Anima 2.9B Preview 1-tier quality, character/artist knowledge and NSFW performance.</p><h2 id="how-it-was-merged">How it was merged</h2><p>Shared blocks <strong>0.8 × Turbo 1.1 + 0.2 × 2.9B P1</strong>; the 12 extension blocks unique to 2.9B P1 are kept as-is. The 0.8 strength was settled by full-scene same-seed comparisons: SFW and 100%-crop details on par across ratios, 0.8 slightly better in NSFW anatomy; no spontaneous mosaic / black bars at any ratio.</p><h2 id="workflow">Workflow</h2><p>UNETLoader (this model) + CLIPLoader (<code>qwen_3_06b_base.safetensors</code>) + VAELoader (<code>qwen_image_vae.safetensors</code>) → two CLIPTextEncode (positive/negative) → KSampler → VAEDecode → SaveImage. No LoRA needed.</p><h2 id="recommended-parameters">Recommended parameters</h2><ul><li><p>steps <strong>8</strong> (8-12)</p></li><li><p>sampler <strong>euler + sgm_uniform</strong></p></li><li><p>cfg: <strong>1.0</strong> without negative; <strong>1.2</strong> with negative (range 1.0-1.6 — negatives take effect as soon as cfg &gt; 1)</p></li><li><p>resolution: 1024×1024 baseline; 1264×848 landscape / 848×1264 portrait tested fine (~14-17 s/image)</p></li></ul><h2 id="prompting-(quality-tags-work)">Prompting (quality tags work)</h2><p>Same as the turbo family — <code>masterpiece, best quality</code> etc. <strong>work</strong>:</p><pre><code>masterpiece, best quality, &lt;safe/sensitive/nsfw/explicit&gt;, &lt;count&gt;, &lt;character (series)&gt;, @&lt;artist&gt;, &lt;appearance tags&gt;, &lt;natural language sentences&gt;</code></pre><ul><li><p>Character name right after the series tag; avoid underscores (<code>hatsune miku</code>, not <code>hatsune_miku</code>)</p></li><li><p><code>@artist</code> strongly steers the style — recommended</p></li><li><p>Write 1-2 full natural-language sentences; for multiple characters describe each one's appearance next to them</p></li><li><p>Standing text-type negative <code>text, letters, sign, logo, watermark, signature, username, artist name</code> + cfg 1.2 recommended: noticeably reduces corner fake signatures / watermarks / logos (useless against text that belongs to the scene itself); for scenes without an artist tag, an empty negative + cfg 1.0 is fine</p></li><li><p>Not good at rendering readable text; for NSFW just write the rating and content tags directly</p></li></ul><h2 id="sample-images">Sample images</h2><p>All samples (s1/s3/s4/s5) are straight outputs of this model at 8 steps, cfg 1.2 with the text-type negative above. Full ComfyUI workflow metadata is embedded in the PNGs — drop into ComfyUI to reproduce.</p><h2 id="related">Related</h2><ul><li><p>Sister project (LoRA version): <strong>Anima29B Turbo v2.9 LoRA (Unofficial Remap)</strong> (<code>anima29b-turbo-v2.9-remap</code>) — same speed on the original 2.9B base without swapping the checkpoint</p></li><li><p><a target="_blank" rel="ugc" href="https://github.com/Langzaigg/ComfyUI-AnimaFlow">ComfyUI-AnimaFlow</a>: AI tool-use API (MCP / HTTP) for Anima-family models, ships a ready-made workflow (<code>fuse</code>) and prompt knowledge for this model</p></li><li><p>Upstreams: <a target="_blank" rel="ugc" href="https://huggingface.co/Gazingstars123/Anima-2.9B">Gazingstars123/Anima-2.9B</a>, <a target="_blank" rel="ugc" href="https://civarchive.com/models/2619830/turbo-for-anima-less-steps">Turbo-ANIMA</a>, <a target="_blank" rel="ugc" href="https://huggingface.co/circlestone-labs/Anima">circlestone-labs/Anima</a></p></li></ul><h2 id="license">License</h2><p>All upstreams (Anima / Anima-2.9B / Turbo-ANIMA) use the <strong>CircleStone Labs Non-Commercial License</strong>; as a derivative work this merged model follows the same license: <strong>non-commercial use only</strong>, and the license notice must be preserved on redistribution.</p><hr /><h1 id="anima29b-p1-+-turbo-1.1-unofficial-merge">Anima29B P1 + Turbo 1.1 合并模型（非官方 Unofficial Merge）</h1><p><strong>非官方（unofficial）社区衍生模型</strong>：由 Anima 2.9B Preview 1 与 Anima Turbo 1.1 经<strong>块级合并（block-merge）</strong>得到的完整 UNET，非 Anima / CircleStone Labs 官方发布。</p><p><strong>直接用</strong>：作为 diffusion model 加载即可，无需任何 LoRA 或补丁。</p><ul><li><p>文件：<code>anima29b-p1-turbo1.1.safetensors</code></p></li><li><p>大小：约 5.4 GiB —— 完整模型（合并版 UNET，40 块 BF16）</p></li></ul><p>目标：turbo 速度（8 步）+ Anima 2.9B Preview 1 档画质、角色/画师知识与 NSFW 表现。</p><h2>合并方式</h2><p>共享块 <strong>0.8 × Turbo 1.1 + 0.2 × 2.9B P1</strong>，2.9B 独有的 12 个扩展块原值保留。0.8 为全场景横评定案：SFW 各配比同档、100% 放大细节同档，NSFW 解剖细节 0.8 略优；任何配比均无自发马赛克/黑条。</p><h2>工作流</h2><p>UNETLoader（本模型）+ CLIPLoader（<code>qwen_3_06b_base.safetensors</code>）+ VAELoader（<code>qwen_image_vae.safetensors</code>）→ 双 CLIPTextEncode（正/负）→ KSampler → VAEDecode → SaveImage。无需 LoRA。</p><h2>推荐参数</h2><ul><li><p>steps <strong>8</strong>（可用 8-12）</p></li><li><p>采样器 <strong>euler + sgm_uniform</strong></p></li><li><p>cfg：不写负面 <strong>1.0</strong>；写负面 <strong>1.2</strong>（范围 1.0-1.6，cfg&gt;1 负面即生效）</p></li><li><p>分辨率：1024×1024 基准，横 1264×848 / 竖 848×1264 实测良好（约 14-17 秒/张）</p></li></ul><h2>提示词（质量词有效）</h2><p>与 turbo 系一致，<code>masterpiece, best quality</code> 等质量词<strong>有效</strong>：</p><pre><code>masterpiece, best quality, &lt;safe/sensitive/nsfw/explicit&gt;, &lt;人数&gt;, &lt;角色 (作品)&gt;, @&lt;画师&gt;, &lt;外貌标签&gt;, &lt;自然语言句子&gt;</code></pre><ul><li><p>角色名紧跟作品标签，标签避免下划线（写 <code>hatsune miku</code> 而非 <code>hatsune_miku</code>）</p></li><li><p><code>@画师</code> 显著影响画风，建议显式给出</p></li><li><p>自然语言写足 1-2 句；多角色逐角色就近描述外貌</p></li><li><p>建议常驻文字类负面 <code>text, letters, sign, logo, watermark, signature, username, artist name</code> + cfg 1.2，可显著减少角落假签名/水印/logo（对画面内容本身的文字无效）；不带画师场景可留空负面 + cfg 1.0</p></li><li><p>模型不擅长生成可读文字；NSFW 直接写分级与内容标签即可</p></li></ul><h2>示例图</h2><p>4 张示例（s1/s3/s4/s5）均为本模型 8 步、cfg 1.2 + 上述文字类负面直出。PNG 内含完整 ComfyUI 工作流，拖入即可复现。</p><h2>关联推荐</h2><ul><li><p>姊妹项目（LoRA 版）：<strong>Anima29B Turbo v2.9 LoRA（非官方块重映射版）</strong>（<code>anima29b-turbo-v2.9-remap</code>）——不想换底模时用，配合原版 Anima-2.9B Preview 1 生效</p></li><li><p><a target="_blank" rel="ugc" href="https://github.com/Langzaigg/ComfyUI-AnimaFlow">ComfyUI-AnimaFlow</a>：Anima 系模型的 AI 工具调用 API（MCP / HTTP），内含本模型对应的开箱工作流（<code>fuse</code>）与提示词知识</p></li><li><p>上游：<a target="_blank" rel="ugc" href="https://huggingface.co/Gazingstars123/Anima-2.9B">Gazingstars123/Anima-2.9B</a>、<a target="_blank" rel="ugc" href="https://civarchive.com/models/2619830/turbo-for-anima-less-steps">Turbo-ANIMA</a>、<a target="_blank" rel="ugc" href="https://huggingface.co/circlestone-labs/Anima">circlestone-labs/Anima</a></p></li></ul><h2 id="license">License</h2><p>上游 Anima / Anima-2.9B / Turbo-ANIMA 均采用 <strong>CircleStone Labs Non-Commercial License</strong>，本合并模型作为衍生作品遵循同一许可：<strong>仅限非商业用途</strong>，再分发需保留许可声明。</p>

Files with SHA256: 0acd73e47996d02cd503dbbcdd30e45d4062db74d23ccef9238785af7e916f37