​

FAQs



Updated on 8th December 2022













​

- What is aipaintr?
aipaintr provides both an API and no code solution for AI image generation. 

Our service offers DreamBooth model training and hosting, which enables the user to generate unique images of any person or subject using Stable Diffusion’s deep learning model.

- How does it work?
Firstly, you upload a set of images of your subject to train your model on.
Next, we take those images and train the model on our servers.
Finally, you can run your directional inputs (called prompts) to generate a batch of new AI-generated images based upon the images you uploaded.

- I'm non-technical - can you explain how it works?
We enable you to generate unique images of  any subject using AI. This could be a person, a pet, a car, a water bottle…anything!
For example, you may want to create some fun and unique profile pictures of yourself for social media.
Refer to the "How to use” resource page for more details.

- Do I need to be able to code to use aipaintr?
No! 

While engineers and developers can use the API, non-technical users can work with our no-code web app. Simply sign up here and you’ll be able to generate AI images in no time.

Step 1: You upload a set of photos of yourself to aipaintr. 
Step 2: We ‘train’ StableDiffusion model to identify what you look like, so you can generate unique images using a text prompt. We will label you with a specific name (e.g. “sks human”).
Step 3: You can then write a prompt - like a short description - of what you want to see. For example "professional headshots of sks human in a pant suit”. This should ideally then produce a unique image of you wearing a suit which you can hopefully use in your profile picture!

- What kinds of images can I produce using aipaintr?
Please head to our Gallery page to see the type of outputs you can expect. 

- What types of images do I need to upload?
You should upload 10 - 30 images of the subject you want to train your model on from various angles and with different backgrounds. Smartphone images work fine! 
In general, the more images you upload, the more representative the output. Refer to the “How to use” resource page for more details.

- Can I download the .ckpt file?
Yes, you can download it once the model training is done.

- How long does it take to train a model?
Depending on the number and size of input images, and requests in queue, it usually takes between 30 mins - 2 hours to train a model.

- What model of Stable Diffusion is aipaintr built on?
aipaintr is built on Stable Diffusion 1.5. We will add additional base models over time.

- How do I buy credits?
To buy credits, you must sign up for an aipaintr account.
Once you are logged in, you can go to billings and purchase credits via a credit or debit card. Stripe is our payment processor.

- Can I buy credits on demand?
Not right now, you have to pay in advance.

- Can I get a discount if I buy credits in bulk?
Please get in touch on support@aipaintr.com

- Can you help me create prompts?
We are actively working on building our own prompt library. For now please feel free to refer to prompt 
# https://stability.ai/sdv2-prompt-book
# https://lexica.art/
# https://prompthero.com/

- Do you offer user training?
Yes, head to our blog post and youtube tutorial channel

- Is my data safe?
Yes. We take data privacy very seriously. You can learn more by reading our Privacy policies.

# Uploaded input images are deleted 24-48 hours (contingent on queue) after model training is complete.
# Your AI generated images are stored on our servers for a 30 days period and deleted after notifying you about the same.
# Your models are hosted securely until you ask for them to be deleted, which you can do at any time.
We are working on additional ability (self-serve) to delete images/models.

- Can I upload kids or minors?
No! It's against our Content Policy. 

- Why are the generated images disfigured or not magical as advertised?
AI can have unpredictable results and may include artistic nudes, erotic or otherwise shocking images. Diffusion models may generate artifacts, inaccuracies and defects in output images - sometimes it is out of our control. So please acknowledge that risk before continuing. 

We can try to minimize its occurrence by following the “how to use” resource page guidelines related to input images. The more variations you get, the better chances for great results!

- Can I get a refund?
Once you upload your photos and we begin the training process, we incur high costs and are unable to offer refunds. Please understand this before agreeing to use our service.

- I have another question!
We’d love to hear from you! Send us a message at support@aipaintr.com and we will get back to you at earliest convenience.