logo
Train your own AI model

IP training

Using the 20 to 80 images you send, AetherAI can automatically train an AI on your IP style.

Example categoriesOther categories can also be discussed depending on your project.
GamesWebtoonsIPLimited series

Prepare 20 to 80 images · About 5 business days · Consultation required

Process overview

This is the step-by-step flow between you and the Aether team, from training to operation.

1
YouStep 1

Reach out

Tell us about the result you want and how you plan to operate it, and the consultation begins.

2
Aether teamStep 2

We review your contact details and continue the consultation

We guide you through the process, required data, and expected schedule in detail.

3
YouStep 3

Send 20 to 80 images

Send the original images that will be used for training to the Aether team.

4
Aether teamStep 4

Receive 10 sample images after training

After training on the provided images, we first share sample images generated by the AI.

5
YouStep 5

Pay for the plan that fits your needs

After checking the sample results, choose the GPU spec and operating model that fit your needs.

6
Aether teamStep 6

Access is granted after payment is confirmed

Once payment is confirmed, we open access so you can use the trained AI right away.

7
YouStep 7

Generate unlimited images with your trained AI

Once training is complete, you can keep generating images in your desired style with the dedicated AI at no extra cost.

Pricing

Choose a shared or dedicated GPU plan based on your workflow and budget.

Easy start

Shared GPU

A shared plan for operating a trained IP model with a reasonable budget.

GPU resources are shared across multiple projects

Well suited to general production and test image generation

Images are generated in request order

Recommended for early adoption, PoC, and small-scale operation

Price

₩350,000 / month

Optimized for operation

Dedicated GPU

A plan that gives you more stable, focused operation with dedicated resources.

Run on your own dedicated GPU

A better environment for frequent use or large-volume generation

Build a faster workflow without queue delays

Recommended for always-on use, large-scale generation, and teams

Price

₩550,000 / month

All plans include a trained IP model. Long-term contracts receive discounts of 5% for 3 months, 10% for 6 months, and 20% for 12 months.

Email or contact information