💬 Discord — <a target="_blank" rel="ugc" href="https://discord.gg/TYHy7fpEXx">join here</a> · ☕ Ko-fi — <a target="_blank" rel="ugc" href="https://ko-fi.com/techannel">support the factory</a> · 💼 <a rel="ugc" href="https://ko-fi.com/techannel/commissions">Request Commission</a>

<h1 id="nangong-yu-zenless-zone-zero">Nangong Yu 南宫羽, 2 Outfits — Zenless Zone Zero 绝区零</h1>

Lead dancer of the Angels of Delusion — halo, wings and all. Now in both stage kits.

<hr />

Character LoRA for Nangong Yu, with a baked identity and a split trigger: every identity tag is absorbed into <code>nangongzzz</code>, so one word carries her face, hair and wings. The outfit token after it picks the kit and the halo. Everything else stays steerable.

<h2 id="versions">🧩 Versions</h2>

<ul>
 <li>🔵 v3.0 Anima (recommended) — split triggers: identity token first, then one outfit token (<code>baseout</code> / <code>museout</code>). Weight 1.0 (0.85–1.0).</li>
 <li>🔵 v2.0 Anima — for Anima / Qwen-Image DiT checkpoints. ✅ Works in the Civitai on-site generator. Both outfits, dual trigger. Weight 1.0.</li>
 <li>🟣 v1.0 Illustrious — for Illustrious / SDXL anime checkpoints. Both outfits, dual trigger. Weight 1.0.</li>
 <li>⚪ v1.0 Anima (legacy) — base outfit only, old trigger <code>nangongyu</code>. Kept for existing gens.</li>
<li>🟠 Remapped Anima 2.9B — this version is the same LoRA remapped onto the 40-layer Anima-2.9B layout (weight 0.5–0.7, no patch node needed). It is a remap, not a retrain, so there may be issues with the 2.9B version — if there is enough interest I may retrain it natively for 2.9B.</li></ul>

<h2 id="outfits-and-triggers">🎽 Outfits &amp; Triggers</h2>

<ul>
 <li><code>nangongzzz</code> — her identity, always first.</li>
 <li><code>baseout</code> — base kit: white frilly dress, red plaid necktie, green bows, and her green halo.</li>
 <li><code>museout</code> — Rhapsody's Muse stage kit: black idol top, frilled black skirt, spiked boots, purple halo.</li>
</ul>

<h2 id="baked-into-the-trigger">🎀 Baked In</h2>

<ul>
 <li>Black twintails with pink inner color and blunt bangs</li>
 <li>Small white wings, spiky halo (green or purple per outfit)</li>
 <li>Red eyes</li>
 <li>The cat hairclip on Muse</li>
</ul>

<h2 id="prompt">📋 Prompt</h2>

Base kit:

<pre><code>nangongzzz, baseout, white dress, green bow, red necktie, detached sleeves, bare shoulders, navel cutout, black choker, black thighhighs</code></pre>

Rhapsody's Muse:

<pre><code>nangongzzz, museout, black shirt, black skirt, chest bow, detached sleeves, bare shoulders, black choker, black nails, brown thighhighs, black boots</code></pre>

Negative:

<pre><code>chibi, speech bubble, english text, multiple views, watermark, signature</code></pre>

Add framing (<code>full body</code>, <code>portrait</code>), pose and scene on top.

<h2 id="prompt-old">📋 Prompt — v2.0 Anima (old triggers)</h2>

For the previous Anima file. One trigger per outfit, no identity token.

Yes, this is the whole thing (Anima):

<pre><code>nangongbasezzz</code></pre>

<h2 id="tips">💡 Tips</h2>

<ul>
 <li>Always give her an outfit token. <code>nangongzzz</code> alone renders her in a neutral, undefined outfit.</li>
 <li>The old words still work on the older versions: <code>nangongbasezzz</code> / <code>nangongmusezzz</code> on v2.0 and v1.0 Illustrious, <code>nangongyu</code> on v1.0 Anima.</li>
 <li>Restyling: keep <code>nangongzzz</code> and prompt other clothes explicitly (<code>casual clothes</code>, <code>swimsuit</code>) — identity, wings and halo hold while the outfit follows your tags.</li>
 <li>Halo color rides the outfit token. If it drifts, lock the other color (<code>purple halo</code> / <code>green halo</code>) in the negative.</li>
 <li>Weight 1.0 (usable 0.85 to 1.0).</li>
</ul>

<hr />

More of my LoRAs: <a target="_blank" rel="ugc" href="https://civarchive.com/user/Techannel">Techannel on Civitai</a> — likes &amp; Buzz keep the factory running ❤️

Nangong Yu 南宫羽, 2 Outfits — Zenless Zone Zero 绝区零 [Anima+Illustrious]

🟠 Remapped Anima 2.9B — this version is the same LoRA remapped onto the 40-layer Anima-2.9B layout (weight 0.5–0.7, no patch node needed). It is a remap, not a retrain, so there may be issues with the 2.9B version — if there is enough interest I may retrain it natively for 2.9B.

Files with SHA256: 3c200b79badfb93a2fb3797d4401fc6c7184de5173ba42025cf997932b73a0fb