[Official] 🎨 Fully Customizable 2D, 3D, and Photorealistic Rendering Styles

Turn your 3D creations into high-quality AI visuals — in 2D, 3D, or realistic styles — with minimal setup and full creative control.

AI Render connects your Character Creator and iClone scenes to ComfyUI, offering a range of professionally tuned styles and easy customization through sliders, text prompts, and reference images.

Whether you’re enhancing realism, aiming for cinematic 3D, or crafting illustrated 2D art, you can fine-tune every result — no node-building or technical setup needed. Check out video tutorials below


:rainbow:22 Built-In Styles — Ready to Use or Customize

AI Render includes 22 professionally tuned presets — 11 for images and 11 for videos — across three core categories:

|676px;x508px;

Realistic

Create photorealistic lighting, cinematic shadows, and lifelike textures. Even lower-quality models can be transformed into high-end renders with added depth and polish.

3D

From game-style CG to painterly looks, 3D presets deliver everything from clean renders to richly detailed, stylized visuals.

2D

Explore anime, hand-drawn, and illustration styles with varied line work, shading, and texture — perfect for creative storytelling and stylized reimaginings.


To keep AI Render accessible for basic hardware setups (see Environment Requirements), it runs locally by default using Stable Diffusion 1.5 for image generation, and Wan2.1 Fun 1.3B Control and Wan2.1 VACE 1.3B for video generation.

:zap:If you’re aiming for ultra-detailed visuals or cinematic results, check out the Advanced Workflow thread for Flux, HiDream, FusionX, and more


:compass: Guide the Style with Prompts & References

Customize what the AI creates using two powerful inputs:

Text Prompts

Use positive prompts to steer the AI toward desired styles or features, and negative prompts to filter out unwanted elements. Prompts are your go-to tool for both styling and fine-tuning detail.

:link: Negative prompts are available by default, but may not be supported in custom workflows using certain third-party models.

Reference Images

Upload any image, and the AI — powered by IP Adapter — will follow its style, lighting, and tone, making it easy to recreate a specific look without needing to write the perfect prompt. This gives you a more intuitive and consistent way to guide your AI results visually.

|960px;x477px;


:control_knobs: Customize “How” AI Works

AI Render gives you full control over sampler behavior — the settings that directly shape how your final image or video looks.

You can adjust key values with simple sliders, offering complete freedom to explore different results. Or, if you want speed, just select from the preset values tuned for each visual style.

What You Can Control:

  • Detail (Steps) – How many iterations the AI uses to generate the result (Higher = more detail)

  • Prompt Influence (CFG)– How closely the AI follows your prompt (Higher = more strict)

  • AI Creativity (Denoise) – How far the final image deviates from your starting input (Lower = more faithful)

All of these can be tweaked easily in the interface — giving beginners smooth entry and experts full creative power.


:hammer_and_wrench: Built for Reallusion Creators

AI Render is fully integrated with Character Creator and iClone, letting you:

  • Instantly stylize characters you’ve already posed or animated

  • Skip manual node-building with beginner-friendly ComfyUI templates

  • Use familiar CC/iC tools like AccuPose, drag-and-drop assets, and real-time scene editing


:rocket: Easy to Start, Powerful to Grow

Whether you want a quick visual concept or a stylized final render, AI Render gets you started fast — and scales with your creativity:

  • Load a preset

  • Drop in your character and motion

  • Tweak your prompt and sampler

  • Generate your styled image or video

:point_down:Video Tutorials to Get You Started

*:sparkles: It’s also the perfect time to try it out — join the Community Challenge and recreate your favorite CC4 character with AI Render for a chance to get Character Creator 5 for free!


:dart: Want more control?

Check out [Consistent Characters and Precise Control for AI Films and Commercial Production]

Learn how to combine 3D ControlNet inputs, apply LoRAs for character consistency, and build custom ComfyUI workflows tailored for high-end AI video production — perfect for creators aiming for cinematic quality and professional results.

3 Likes

THIS!!..is what I been waiting for!!

2 Likes

yeah +1

finally a place to discuss 3D in combination with AI !

1 Like

I have been lingering on Iclone 7/ CC3 waiting for a compelling reason to upgrade.
and this is it I was planning to go “All in” on an AI workflow and buying all new hardware, when I fully retire in 2026
This is the impetus I was seeking
Congratulations Reallusion. :+1: :+1:

2 Likes

this have huge consistency issues , this cant be used for professional jobs …

Are you beta testing this?

Looks like they have the options to train Loras to get better consistency. See here: https://discussions.reallusion.com/t/official-consistent-characters-and-precise-control-for-ai-films-and-commercial-production/

Aside from that, you also need to check if you can use certain AI image or video generation models for commercial usage. It’s not possible with all.

1 Like

so i will have to train every single character, prop and environment each time i want to create a scene? every single character must be trained even if the scene has 10 characters?

  • if you want that the AI knows how your character looks like from all sides, you have to train.
  • No, it is enough to train your hero model/models. Others in the background can be prompted…
  • You can use an image for the environment and use the canny and depth controlnets for the geo
  • mostly it is enough to use images
  • and use “fixed” for generating

… and I know x studios using AI for professional jobs.

It is just the beginning, AI development is very strong, every 2 weeks new updates. Imagine what’s possible in 1 year.

3 Likes

Not necessarily. It’s possible to use IP-Adapters (Image prompts) to transfer image styles from a reference image into the generated images. This also allows to transfer faces. Properly trained Loras should give better results. Maybe the image prompt input mentioned above can help in this situation.

It’s also possible to use a face swapper, for example Face Fusion. But here is the question how well will it work with cartoon style or artistic styles which are not photorealistic.

1 Like

I converted my 3D character to a realistic image…

Honestly, I’m still struggling to find an advantage… But I’ll keep trying.


1 Like

Same here. This is indeed very exciting! The nay-sayers won’t have to use it.

1 Like

Unfortunately, I can not join the contest because I’m in Canada for three months and only have a lowly laptop with a Nvidia 1070. :sob:

AI-enhanced movies with AI voices! Yeah! Onwards and upwards!

1 Like

It doesnt apply the prompts. Or I’m doing something wrong. Because I write like in the tut video – in an urban environment-- and it returns with a ‘solid’ background

If iClone’s new AI renderer can already crank out super realistic, photogenic results, what’s the point of purchasing CC5?
I’ve been trying to wrap my head around it, but it’s been all brick walls so far.

play with these settings :

1 Like

It’s mainly focused at the Unreal MetaHuman customers, which is probably the right thing to do. If you don’t create and animate and render characters in Unreal Engine, there isn’t much new. It looks like CC5 is streamlining the process to create customized MetaHuman characters a lot.

For one, CC5 provides much better facial expressions, so that would still be valuable. The AI solution is focussed on rendering, but I don’t think that it will enhance the characters in other ways.

It’s like Reallusion is saying: “If you’re building for Unreal, we’ve got your pipeline covered.”
I wonder if they’ll eventually offer more for non-Unreal users. Would be great to see CC5 flex a bit more