Unlocking AI with SD3

Stable Diffusion has been a game-changer in AI, allowing users to generate stunning images from simple text prompts. With each update, these models have grown increasingly powerful, and the latest version, SD3, has everyone excited. However, there’s a downside—Civit AI has banned the use of SD3 due to stability concerns.

But don’t worry! This blog will introduce you to an incredible new AI image generator that lets you access and utilise all SD3 models and checkpoints, so you can still enjoy the full potential of this groundbreaking technology.

For better understanding see the video blog here: https://youtu.be/iatGSdNNwY4?si=YKceIf-ZBKPTqKbA

Shakker AI

Introducing Shakker AI a powerful AI image generator and models hub for art lovers. With this amazing tool, you can create any kind of image with this tool.

• So first, create an account with your Google or email account once you are logged in you will find a vast number of image generation models for various use cases.  For example, Photography, illustration, architecture and space design, game design, anime, 3D and many more.
• To find your desired model you can set the filter from here for example if you want to use only models which are based on the new SD3 select it. From the top of the base model, you will find the model type select checkpoint there.

• You can also choose the Laura and other models later in the image generation page.
• On the bottom you will find the model application scope and commercial license. If you select generated images commercially usable, then you will get the commercial license of your generated image.

• Now the picture given below indicates your filtered models from where you can select any checkpoint model to generate the images for example if you want to generate images with  SD3 just select it.

• You will get the model page first where you will find three options- 1) Run the model, 2) Add to favourite, and 3) Download.

Yes, you can download the model and install it locally. Now if we click on the Run model, it will redirect you to the image generator page. Another way to access the page is to click on this image generator option from the homepage. On the left side of the image generator interface, you will notice some images generated by Shakker AI. To develop your pictures, you have to come to the right side and on the top, you will find the prompt box.

• So, first add the text prompt in the box.
• Then under The Prompt box you will find the negative prompt option to set which means what you don’t want in your image.
• Then you will find some negative prompt presets also available there.
• Next is the base model checkpoint which you have already chosen, however you can change it.
• Then click on the model name and you will find all the checkpoint models from Shakker AI before you can set the filter to find the desired one.
• After selecting the base model, come to the enhanced model or Laura option. To add any Laura model, click on ADD Laura and you will find a vast number of Laura models for various styles. Select your desired one you can multiply Laura to get more personalised output.
• Then you will find the options for image reference. If you want to generate a new image with any image reference, then you can add it.
• Next in the parameter settings, you will find some Advanced options too.

• Same with sampling steps where you have a default 20 but if you want more can go ahead and add in more.
• Next the seed if you want to use a fixed set to generate consistent images you can enter the seeds.
• Next you will find a very important feature – Fix hand and Face. Sometimes various AI generate unusual hand or facial features so this option is to fix them.
• After that you can make some changes down here, particularly image size. It defaults to 1024 x to 768 but you can use a variety of other sizes.
• Then if you look at the image number, it’s set to two and you can go ahead and Jack that up to four. So, if you do that, it would use 52 credits to 13 credits per image but if you use any model from XD l or XD 1.5, it will cost only two credits per image.
• Now hit the generate button and get four images.

Now you will find some options under every image-

• First the very stable, if you want to generate a variation of any particular image click here and you can see it generates four variations of the first image.
• Next if you want to reuse any image prompt but can’t remember the prompt, simply click on the prompt and the prompt using the image will be added to the prompt box.
• Next if you want to find the model you use to generate any image, click on the match model button.
• In the end if you want to use any image as the reference image you can do it by simply clicking on the reference button. okay now I’m going to put that

Now you can rearrange the images as you see fit if you are a comic artist or if you create comic books or storyboarding and you want to make sure that your images are in order. So now you have got your images. Let’s take a look at this and click on the download icon to get it. Now open it and check this out and you may notice there’s no watermark and it will allow you to try it for free.

Some Other Features

On the top of the canvas, you will find some other features to customise your generation.

• First of all, the Inpaint option with which tool you can paint any part of your image for example if you select an image and want to change the smaller character into another object select the Inpaint tool first.
• Then you can select an object with the auto selector tool and rectangle selector.
• After selecting the area come to the Prompt box and enter your desired object description.
• After entering the description hit the generate button and you can see ShackkerAI generates four variations for you and you can select anyone and send it to the canvas.

Upscale Image

With this tool, you can upscale any image up to 4X following the methods-

• First, select the image and click on the upscale image option.
• Now set the upscale value, parameter scale, and style.
• Next in the advanced option, you will find options like Details, HDR, Similarity, and Prompt.
• Once everything is set hit the generate button and get the upscaled image ready.

Remove Background

With this feature, you can remove any image background in a single click. Also, with this removal tool, you can remove any object from the image.

• First, select the image and then select the area with the rectangle selector tool.
• Now hit the remove button.
• After that, you can crop any image with the crop tool.

Next with the Outpaint tool you can expand any image with it. If you want to make a collage then select all the images you want to use and then hit the collage button and get your collage ready. In this method, you can create and customize images with this AI. Most importantly, you will get 200 credits per day for free if you need more than you can buy one of the premium plans mentioned below-

• The first plan starts from $8 which is the basic plan and you will get unlimited access to member-exclusive models which allows you to use generated pictures for commercial purposes with 10,000 tokens.
• And $20 for an advance plan with 30,000 tokens.

Some FAQs

Q1: What is the usability of the feature of the commercially usable image?

By selecting commercially usable images in the ShakkerAI tool, you gain a commercial license for your generated images. This license allows you to use the images in various commercial projects, such as marketing, product design, or content creation, without worrying about legal restrictions. It provides flexibility and the freedom to monetize your AI-generated content.

Q2: How do get fine-tune large-scale models?

Just come to the enhanced model or Laura option and get on it. Laura’s low-rank adaptation is a method designed to fine-tune large-scale models more efficiently.