Candy AI Image Generator doesn’t feel like a typical “prompt in, image out” generator. It feels more like directing a scene, than creating one.
You pick a character (game changer!) not a blank canvas, but a personality. Then you can adjust the character’s clothes, pose, and overall aesthetic. Then, somewhere between the adjustments, you stop thinking, “How does this software work?” and start feeling like… you’re co-creating with someone?
Have you ever been in that headspace where you’re trying to conjure up a scene in your mind, and it just doesn’t seem to fit? Candy AI Image Generator fills that need. Not perfectly, but enough to get you hooked.
How to use Candy AI Image Generator: Step-by-step guide
This generator is quite easy and follows a three-part procedure:
- Select a Character
- Make the prompt, and generate
- Sign up
With that ordering, you are able to go through it, choose a character, configure your image, and then create an account when the service prompts you for this.
Step 1: Choose a Character
The initial screen you come across after clicking Generate AI Images is the character choosing page:
Generate Image Choose character
This is where you will make the decision about which character you want to have in your generated image.
Main features on this screen
| Feature | What it does |
| Anime tab | Shows anime-style characters |
| Guys tab | Switches to male characters |
| Character cards | Displays available AI characters |
| Highlighted border | Shows which character is selected |
| Select button | Confirms your character choice |
Character cards
You can see a few character cards, including:
- Nari
- Yama
- Trixie
Every card features a character preview picture and their name. Yama is selected, as you can see it has a bright pink border around the image.
Select button
A large, gradient button labeled Select is located at the bottom. Once you’ve selected your character, click Select to be taken to the image creation screen.
How to complete Step 1
- Go through the grid to find a character.
- You can use Anime or Guys to toggle between categories.
- Select the character you want.
- Double-check that the card for the selected character is highlighted.
- Click Select.
Step 2: Write a Prompt and Generate the Image
Once you’ve chosen the character, you will be presented with the image generation screen. Here, you can give the AI instruction on what to generate. At the top, you’ll find the text, Generate Image. To the left of this, you will see an option for the character you chose (in the screenshot provided, this is Yama Gito).
Main features on this screen
| Feature | What it does |
| Character preview | Shows the selected character |
| Prompt box | Let’s you describe the image |
| Suggestion tabs | Helps build prompts faster |
| Suggestion chips | Adds quick prompt details |
| Number of images | Chooses how many outputs to generate |
| Generate Image button | Starts the generation process |
Character Preview
On the left, you can see which character you’ve selected. This is to ensure you’re not accidentally generating an image of the wrong character. That costs both time and credits! There’s also a small, circular button on the character card (which isn’t explained in the screenshot). It looks like something related to characters, most likely a button that resets or cycles through characters.
Prompt Box
This big text area is where you type the description of the image you want to generate. This is the example prompt that was visible on our screenshot. It’s the description for:
- a scene
- a character’s clothing
- a body position/pose
- an expression/gaze
A good prompt should be detailed and clear.
Prompt structure example
| Prompt part | Example |
| Setting | sitting on a leather sofa |
| Outfit | wearing a fur jacket |
| Pose | seated, relaxed |
| Expression | looking at the viewer |
| Mood | confident, cinematic, playful |
An example of a basic text prompt:
“Yama sitting on a leather sofa, wearing a stylish outfit, looking at the viewer, soft lighting.”
Suggestion tabs
There are tabs below the prompt area:
- Suggestions
- Outfit
- Action
- Pose
- Accessories
- Scene
They let you specify certain details, so you can use your own prompts rather than creating a prompt for everything from scratch.
Suggestion chips
And below the tabs, there are several chips you can choose from such as:
- Bikini
- Skirt
- Lingerie
- Crop top
- Leather
- Mini skirt
Clicking on any of these probably will add this detail to the prompt you are typing or will change the prompt in some way.
Number of images
You can choose to generate 1, 4, 16 or 32 images. 1 is what we’ve selected here, as you can see from the screenshot. Some of the more options have a few little icons, which likely indicates that they require more of your credits (or perhaps you will need to upgrade your plan).
Which number should you choose?
| Option | Best for |
| 1 | First test, fastest and safest |
| 4 | Comparing a few variations |
| 16 | Exploring many results |
| 32 | Large batch generation |
Start with 1 for the first try. When you are satisfied with the generated output, then move to a higher value to generate more results.
At the bottom of the page, you’ll find:
Generate Image
Hit this when you have configured your prompt and batch size.
Complete Step 2
- Choose the character from the preview
- Type the prompt into the prompt box. You can use the tabs on this box such as Outfit, Action, Pose, Accessories and Scene if you require more prompt suggestions
- Choose the batch size
- Press the Generate Image button
Step 3: Create account
After you generated or kept going on with it, the platform then showed you with an account screen.
Here’s where you can choose to store your selections
What you can find on this screen
The screen has 2 different parts:
| Section | What it shows |
| Left side | Character preview and Candy.ai branding |
| Right side | Account creation form |
On the left, you see a message that says, “Your choices will be saved.” You can deduce that when you create an account, the characters you choose as well as your preferences, and the history of content you generate, are all stored with your account.
Main features on the account screen
| Feature | What it does |
| E-mail field | Enter your email address |
| Password field | Create a password |
| Minimum 6 characters | Shows password requirement |
| Create Free Account | Registers your account |
| Sign up or log in with Google | |
| Discord | Sign up or log in with Discord |
| X | Sign up or log in with X |
| Sign in | For users who already have an account |
| Terms of Service | Legal agreement link |
Email and password fields
The arrows are pointing to the two major fields:
- Password
The password field comes with the note: Minimum 6 characters
Social login options
If you don’t want to create another password, you can proceed with:
- Discord
- X
Create Free Account button
After entering your details, click: Create Free Account
Completing Step 3
Enter your email Enter a password consisting of at least 6 characters. Click Create Free Account. Alternatively, you may use Google, Discord, or X for a quicker registration process. If you have an account already, click Sign in.
Full Workflow Summary
| Step | Action |
| 1 | Choose a character from the character grid |
| 2 | Write a prompt, choose image count, and click Generate Image |
| 3 | Create an account or sign in to save/continue |
Tips for getting better results
- Start with 1 image First, try just one image. You can see if your prompt is good at getting results.
- Use the prompt helpers Tabs and chips are here to assist you. If you don’t know where to begin, use Outfit, Pose, or Scene.
- Keep it straightforward In short, a specific, succinct prompt works best.
- Stick with 1 character Keep using the same character to get better consistent results.
- Generate more after testing Once you have a good prompt, click on 4, 16, or 32 to get more results.
Wrapping Up
Overall, the process is really well thought-out:
- pick a character
- describe the picture
- generate
- make an account
The best features are the clear character cards, prompt help, photo selector, and account setup. While it is very intuitive, it also allows those who desire to tweak the final results with more detailed text prompt options.
