<h1 id="pearly-desire">♦♦Pearly Desire♦♦</h1>An experimental merge of Opal and Anima mix, This one tries to bring a more detailed result while still leaning towards a more comic like feel especially in the clothing. Long detailed prompts get the best results,<h1 id="installing-for-local-gen">♦♦Installing for local gen♦♦</h1>Get the text encoder and VAE from the <a target="_blank" rel="ugc" href="https://huggingface.co/circlestone-labs/Anima">HuggingFace page</a>.The model is natively supported in ComfyUI. The model files go in their respective folders inside your model directory:<ul><li>Checkpoint goes in ComfyUI/models/diffusion_models</li><li>qwen_3_06b_base.safetensors goes in ComfyUI/models/text_encoders</li><li>qwen_image_vae.safetensors goes in ComfyUI/models/vae (this is the Qwen-Image VAE, you might already have it)</li></ul><h1 id="settings">♦♦Settings♦♦</h1><h3 id="settings-that-i-use-with-this-checkpoint">Settings that i use with this checkpoint</h3>Note: I’m pretty much winging it.1st Ksampler<ul><li>Sampler: er_sde</li><li>Scheduler: Simple</li><li>CFG: 5 ish</li><li>Steps: 20 - 35</li></ul>Ultimate SD Upscaler<ul><li>Sampler: er_sde</li><li>Scheduler: Simple</li><li>CFG: 5 ish</li><li>Steps: 5 - 20</li><li>Tile size: 512 or 1024</li></ul><h3 id="recommended-positives:">Recommended positives:</h3><pre><code>(masterpiece, best quality, good quality, score_9, score_8),</code></pre>Recommended negatives:<pre><code>(lowres, bad quality, worst quality, long neck, bad anatomy, watermark, jpeg artifacts, ugly, score_1, score_2, score_3), , </code></pre>Additional notes:This model performs best with more detailed prompts. Dont forget you can also use weights on prompts example: (standing:1.2),

Files with SHA256: ec98383b42c047c80cf6cd8e5d0751c9e6ea8c17ddebb60e38b8f124d0473e03