Generating Images, Icons and Backgrounds
When the stock libraries don’t have what you need, Glintia can generate it. Describe the picture and choose a model, and the result lands in your design. One generator works in three ways:
- AI Image for any picture, prop or illustration, and for editing a picture you already have
- AI Background for a full-scene background shaped to your canvas
- AI Icon for square icons, with a see-through background or as an SVG
All three use AI credits. The price is shown before you generate.
Open the image tools
Section titled “Open the image tools”| Tool | Where to find it |
|---|---|
| AI Image | AI Studio › Image › Generate an image (or Edit with AI). Also the Generate Image button at the top of the Image / vector panel |
| AI Background | AI Studio › Image › Backgrounds. Also Generate AI Background in the Background panel |
| AI Icon | AI Studio › Image › Icons. Also Generate Icon at the top of the Elements panel |
In AI Builder mode, AI Image, AI Background and AI Icon each have their own button on the rail.
Generate an image
Section titled “Generate an image”- Open AI Image. The panel opens on the Create tab
- Under Model, keep the default (Nano Banana 2.1) or click it to choose another model (see Choose a model)
- Under Prompt, describe the picture: subject, setting, light and mood. Click Idea for an example prompt if you want a starting point
- Optionally pick a Style, such as Photo, Illustration, Flat vector or Watercolor
- Pick a Shape. Slide (the default) matches your canvas, or choose an aspect ratio such as 1:1, 9:16 or 16:9
- If the model offers them, set Resolution and Quality
- Use the − and + buttons next to the Generate button to choose how many images to make
- Click Generate (for example Generate 2 images), or press Ctrl+Enter (Cmd+Enter on a Mac) in the prompt box
The images generate in the background, so you can keep editing. Open the My images tab to watch them arrive (the tab shows a pulsing dot while anything is being made), then click a finished image to add it to the current scene.
Image settings
Section titled “Image settings”- Style: None, Photo, Cinematic, Illustration, Flat vector, 3D character, 3D render, Anime, Comic, Watercolor, Oil paint, Sketch, Pixel art, Isometric, Clay, Product and Retro wave. Some options fold behind a +N chip when the panel is narrow
- Shape: Slide matches your canvas’s proportions. Each model lists the aspect ratios it supports
- Resolution: 1K or 2K, and 4K on some models. This is the size of the generated picture only. Your video’s export quality still depends on your plan
- Quality: Draft, Standard or Best, on models that offer it. Higher quality costs more
- Transparent background: on models that support it (GPT Image 2.5 and GPT Image 2.5 Sunburst), for stickers, props and cut-outs
- How many: up to 4 images per run on most models
Hover over a Resolution or Quality option to see its price per image. The credit total above the Generate button updates as you change things.
Edit a picture with AI
Section titled “Edit a picture with AI”You can give the generator pictures to work from: a photo to change, or a person, product or style to keep.
- Select an image on the canvas
- In AI Studio › Image, click Edit with AI. AI Image opens with your picture added under References
- In the prompt box, say what to change or create, for example “Make it night time with warm window lights” or “Put this product on a marble kitchen counter”
- Click Generate
You can also add references yourself in the References row. Click the pointer icon to use the image selected on the canvas, or + to choose from your files or upload a new picture. The hint next to the row shows how many references the current model accepts.
On a finished image, hover and click Use as a reference to feed it back into the next run. If the current model can’t take references, Glintia switches to Nano Banana 2.1 and tells you.
Generate a background
Section titled “Generate a background”AI Background makes full-scene backgrounds shaped to fit your canvas.
- Open AI Background
- Describe the scene behind your slide: the place, the light, the mood, and where to leave space for text
- Pick a style: None, Natural, Landscape, Dreamy, Soft blur, Pastel, Illustrated, Flat vector, Watercolor, Paper, Retro grid, Holographic, Iridescent, Chrome, Cyberpunk or Soft float
- Click Generate (2 backgrounds by default)
- Click a finished background, or hover and click Set, to make it the current scene’s background
The shape follows your canvas (the closest shape the model supports), and backgrounds start at 2K resolution so they stay sharp full-screen. Setting a new background replaces the one already on the scene.
Generate an icon
Section titled “Generate an icon”AI Icon makes square icons, sized like icons when you add them rather than filling the scene.
- Open AI Icon
- Say what the icon should show, for example “rocket launch” or “secure padlock”
- Pick a style: Flat (the default), Solid, Outline, Colourful, Sticker, 3D, Glossy, Pixel or Hand-drawn
- Click Generate (4 icons by default)
- Click an icon to add it to the scene
The default model for icons, GPT Image 2.5, has Transparent background switched on, so icons come out with no background to remove.
Choose Recraft V4.1 Vector to get real SVG files that stay sharp at any size. It makes vector-style art only, with no photos or soft lighting.
If you pick a model that can do neither, the panel reminds you which models can.
Choose a model
Section titled “Choose a model”AI Image has 20 models from providers such as Google, OpenAI, ByteDance, Black Forest Labs, Ideogram and Recraft. Click the model box to open the picker:
- Recommended, Best value and All models sections when you’re browsing
- a search box, and Sort by Recommended, Cheapest first, Newest first or Fastest first
- Best for filters, such as Photoreal, Text & logos, Product shots, Consistent characters or Fast drafts
- Must have filters: References, 2K+, Transparent and SVG
- each model’s typical price and wait time, and a price tier from $ to $$$$
- a star to add a model to your Favourites, and an info button that shows what the model is good at, what to watch out for, and its specs (shapes, resolutions, quality levels and how many references it takes)
Glintia remembers the model you last used in each of the three tools, in the browser you’re using.
Credits
Section titled “Credits”Each image’s price depends on the model and your settings. The picker shows a model’s price for one image at default settings, and the bar above the Generate button shows the exact total for your run. Some examples:
| Model | Credits per image |
|---|---|
| Nano Banana 2.1 (default for images and backgrounds) | 5 at 1K, 6.75 at 2K, 14.25 at 4K |
| GPT Image 2.5 (default for icons) | 0.75 Draft, 1.5 Standard, 5.5 Best |
| Recraft V4.1 Vector (SVG) | 6.75 |
| FLUX.3 Klein | 0.75 |
| Z-Image Turbo | 0.5 |
| Nano Banana Pro | 12.5, or 25 at 4K |
Reference pictures add to the price on some models. With the default settings, a run of 2 images costs 10 credits, 2 backgrounds cost 13.5 credits (they start at 2K), and 4 icons cost 6 credits.
If an image fails, its tile says Credits refunded and offers Retry.
Work with your results
Section titled “Work with your results”Open My images (or My backgrounds or My icons) to see everything you’ve made with that tool. Hover over an image for these buttons:
- Add (or Set for a background) puts it in your design
- Turn into a video opens the AI Video generator with this picture as the starting frame. See AI Video, AI Actor and Avatars
- Use as a reference feeds it into your next run
- Another like this runs the same settings again (the button shows the cost)
- Open full size opens the image in a new tab
Every result is also saved to your files, so you’ll find it in the Upload panel later.
Remove a background with AI
Section titled “Remove a background with AI”To cut the subject out of a picture you already have, select the image and choose AI Studio › Image › Remove background. It costs 0.25 credits per image. See Removing Backgrounds for the details.