It's no secret that I love experimenting with my prompts, trying out different styles and artists without relying on any additional LoRAs. I can safely say that the Anima model truly lets you feel the magic of "prompt alchemy," opening up a massive amount of creative possibilities.

⚙️ Recommended Settings

• Steps: 20-40 (Best around 30)

• Sampler: Euler

• CFG Scale: 3 to 5

• VAE: Already integrated (Baked-in)

✨ Key Features & Strengths ✨

✅ Natural Language & Tag Mastery

It understands natural language just as well as Danbooru tags. You can describe a scene with plain words, use standard tags, or mix both. This is one of the first anime models where this feature works exceptionally well.

You don't need to be a prompt engineering expert; the model will understand you anyway!

✅ Massive Built-in Visual Styles

A single checkpoint effortlessly handles a huge variety of visual concepts. It can draw classic anime, detailed illustrations, soft pastel art, and stylized digital art.

As always, it all depends on what you write in your prompt. Just ask for it, and the model will understand!

✅ Excellent Multi-Character Composition

It confidently builds scenes with multiple characters using direct spatial instructions like "girl on the left / on the right".

This is exactly what I was missing in other models. Being able to properly position and arrange characters was a huge struggle for me, especially when creating art for my OCs, Aelis and Reito. Now, I can finally breathe easy

💡 Tips & Resources 💡

🎨 Artist References:

To reference specific art styles, use the @artist name format in your prompts. For example: @nixeu, @sheya, @nat the lich - the model will recognize and apply their distinctive styles.

📚 Style Guide:

For inspiration and examples of various styles you can achieve with this model, check out my first style guide article:

👉 Style Exploration Guide

🔜 Coming soon: An updated style guide specifically created for Melys Anima with detailed examples and prompt templates!

⭐Who is Melys Anima for? ⭐

Everyone! From prompt gurus to absolute beginners.

If you are just starting your journey in AI art generation, don't want to memorize Danbooru tags, prefer writing prompts in plain language, and want a model that truly understands you and generates the exact result you want - even from long and complex scene descriptions - you are in the right place!

Enjoy the magic of Melys Anima! 🎨✨