Guide to getimg.ai Generators and Actions
The nine generators and ten Actions in getimg.ai, what each one does, and how an Action page works from source file to finished result.
Generators and Actions
getimg.ai has two kinds of tools, and knowing which one you want saves the most time.
- Generators are the nine general-purpose tools in the sidebar. Each one covers a whole job rather than a single outcome, so you steer it yourself. Five start from a prompt you write; the other four start from a file you add.
- Actions are presets, each built to produce one specific result. You add a source file, pick from fixed menus, and generate. There is no prompt to craft, though a few ask for a short detail in plain words.
Both run on the same account, plan, credits and models, and neither overwrites your source: every run produces a new file and leaves the original in your gallery.
The generators
The sidebar groups the nine generators under Image, Video and Audio. Click a group to open it, then pick a tool.
- Image: Create image, Resize image, Remove background, Upscale image
- Video: Create video, Upscale video
- Audio: Create music, Create speech, Create sound effect
They do not all work the same way. Create image, Create video, Create music, Create speech and Create sound effect start from a prompt box at the bottom of the page. Resize image, Remove background, Upscale image and Upscale video have no prompt box at all: you click the box at the bottom to add a file, or drag one in, then set the options that apply, such as the target aspect ratio or resolution.
What every generator page does share is its layout: a gallery above showing everything you have made with that tool, and the controls below, which are a prompt box, an upload area, or both. Click the arrow to start. Originals are never overwritten, so each run adds a new file and leaves your source in place.
Each generator has its own walkthrough: creating images, creating videos, generating music and generating sound effects.
How an Action page works
Every Action follows the same four steps, so once you have run one you can run any of them.
- Open the Action you want. Each one lives on its own page, at a URL under /app/actions/.
- Add your source file in the box on the left: upload it, drag and drop it, or pick one straight from your gallery. Character creator is the only live Action that needs no file.
- Choose from the options listed on the page. Most are fixed menus rather than free text. A few Actions add an optional field for what the menus do not cover, and UGC video asks you to type the product name and its selling points, because the presenter has to say them out loud.
- Set the output options that Action offers, such as quality, aspect ratio, clip length or how many versions you want, then click the arrow. Results appear in the gallery on the right.
Not every Action offers a version count. Where one does, it is a + / − stepper that starts at 1, so you can ask for any number up to that Action's maximum rather than picking from fixed values. Sticker maker is the exception: it makes one sticker per style you select, so the number of styles is the number of stickers.
The Actions
Image Actions (6)
- Headshots: Professional headshots from a single selfie. Input: a clear, well-lit photo of the face looking at the camera. Options: outfit (keep what you are wearing, suit, blazer, shirt, knitwear, t-shirt, leather), backdrop (white, studio grey, office, office with a skyline, data centre, street, nature, sky, black, bookshelf, stage), treatment (corporate, creative, startup, natural, editorial), aspect ratio 4:5, 1:1, 3:4 or 16:9, quality 2K or 4K, up to 8 headshots per run.
- Photo restoration: Repairs scratches, tears, blur and fading in an old photo. Input: the damaged photo, scanned as sharply as you can. Options: restore only, which keeps the original colours, or restore and colorize, which adds colour to a black-and-white original. Quality 1K, 2K or 4K.
- Background cleanup: Removes unwanted people and objects from the background while the subject and the scene stay put. Input: the photo with the distractions in it. Options: quality 1K, 2K or 4K. For a transparent cut-out instead of a cleaned scene, use the Remove background generator.
- Relight: Changes the lighting of a photo without changing anything in it. Input: the photo to relight. Options: the direction the light comes from, its colour, and intensity from 1 (subtle) to 10 (strong). Quality 1K or 2K.
- Sticker maker: Turns one subject into die-cut stickers on a transparent background. Input: one subject, close up and well lit. Options: up to 6 styles at once from kawaii, bold outline, chibi, shoujo, retro anime, 3D cartoon, vinyl toy, glossy emoji, 3D clay, pixel art and watercolor. You get one sticker per style, so the style count is the variation count. Quality 1K or 2K.
- Character creator: Builds an original character portrait from a set of traits, RPG-creator style. Input: nothing. This is the one image Action that needs no source file. Options: gender, heritage (up to 3 blended), build, skin tone, skin detail, hair style, hair colour, eyes, age, outfit and art style (photoreal, anime, 3D render, comic, painterly), plus a free-text field for anything else. Aspect ratio 3:4, 1:1, 4:5 or 9:16, quality 2K or 4K, up to 4 portraits per run.
Video Actions (4)
- UGC video: A presenter talks to camera about your product, in the style of a creator review. Input: a head-and-shoulders photo of the presenter, 1 to 3 product photos, the product name as it should be spoken, and what makes the product good. Options: tone (casual, excited, review, explainer, recommending), length 5, 10, 15, 20 or 25 seconds, quality 480p, 720p or 1080p. This one takes several minutes to render.
- Product orbit: A 360° clip with the camera orbiting your product on a clean backdrop. Input: 1 to 3 photos of the product. Options: camera move (level orbit, rising, high angle, push in, low hero sweep, pull back), backdrop colour, aspect ratio 16:9, 1:1, 9:16 or 4:3, length 5, 8 or 10 seconds, quality 720p or 1080p, plus a free-text field for setting, surface or lighting notes.
- Dolly zoom: The vertigo effect: the subject holds still while the world stretches away or closes in. Input: one photo for the camera to move through. Options: motion (expand, collapse, or whip, which goes there and back), strength 1 to 5, length 5, 6 or 10 seconds, aspect ratio original, 16:9, 9:16 or 1:1, quality 480p, 720p or 1080p.
- Image transition: A clip that travels from one image to another. Input: a start image and an end image. Options: transition type (morph, camera travel, zoom through, dissolve, whip pan, or a flash of light), length 4, 5 or 8 seconds, aspect ratio original, 16:9, 9:16 or 1:1, quality 480p, 720p or 1080p.
Actions do not add new models. They package the same models the generators already use behind a fixed, one-purpose flow, so your plan, your credits and your commercial rights work exactly the same way whichever route you take.
Frequently Asked Questions
Generators are the nine general-purpose tools in the sidebar: Create image, Resize image, Remove background, Upscale image, Create video, Upscale video, Create music, Create speech and Create sound effect. Each covers a whole job and you steer it yourself. Five start from a prompt; the four reshaping and upscaling tools have no prompt box and start from a file you add. An Action is a preset built for one specific result, such as Headshots or Product orbit: you add a source file and pick from fixed menus. Both run on the same account, plan, credits and models.
No. Actions run on the same plan and the same credit balance as the generators, and they use the same underlying models. What you pay depends on what you generate, for example the quality you pick and how many versions you ask for, not on whether you used an Action or a generator to do it.
There is no prompt to craft: the options are fixed menus you pick from. Some Actions do take a little typed input. Character creator, Product orbit, Dolly zoom and Image transition each add one optional free-text field you can leave empty, and UGC video requires two short entries, the product name and what makes it good, because the presenter speaks them in the video.
Four of them: UGC video, which has a presenter talk to camera about your product; Product orbit, which turns product photos into a 360° clip; Dolly zoom, which applies the vertigo effect to a photo; and Image transition, which travels from one image to another. The other six Actions all output images.
No. Every Action produces a new file and leaves your source untouched in the gallery, the same way the generators do. You can run the same source through an Action as many times with different options as you like and compare the results side by side.