Short answer: Use the image creation flow to turn an idea into an AI-generated image. You can write the prompt yourself, or select create from image to have Typeshutter draft an editable prompt from one image.
Steps
Sign in to Typeshutter.
Open Create and choose Image. You can also start from the Create composer that stays in view on the homepage; submitting there opens Create, where the new card appears and keeps updating.
In the prompt box, describe your shot in text. If you want help starting from an existing image, select create from image.
If you use create from image, choose one image. Typeshutter analyzes that image and fills the main prompt field with a generated prompt you can review and edit.
Review the prompt before creating. The helper does not submit the generation and does not spend credits by itself.
If you want the generation to use separate image references, add them with the normal Image control. The create from image helper image is only used to write the prompt and is not automatically attached as a generation reference.
Open Settings to choose the model, aspect ratio, and resolution.
Select Create and wait for the result to appear in your library.
If you are signed out, you can still type a prompt in the homepage composer. Submitting sends you to Log in, and your prompt is restored on Create afterward so you can review and submit it. Adding references and create from image require you to be signed in.
Create several related images
Include a number in your prompt to get a set, such as "Create a 7-slide Instagram carousel..." or "Generate 8 product photos...".
Typeshutter creates 2 to 10 images as separate assets in one post. Requests for more than 10 images are rejected before generation.
The prompt must include a number. "Create an Instagram carousel" without a number still creates one image.
Each image is charged at the selected model's per-image price, so a 7-slide carousel costs seven times one image.
The first image sets the shared visual style and is used as a reference for the rest.
Multiple videos in one prompt are not supported. Create one video at a time.
Guided AI tools
For a focused job, choose a tool under Tools in the Create picker, or browse them all at typeshutter.com/tools. Each tool takes one photo and an optional short direction. See Use Typeshutter AI Tools for what each tool keeps, changes, and returns.
Professional retouch: remove temporary flaws from portraits, wedding photos, and group photos while keeping identity, skin tone, and real skin texture.
Golden hour 35mm: re-render a photo with a clean, visible 35mm film look.
Profile picture: turn one photo into a square, avatar-ready head-and-shoulders picture.
AI portrait generator: turn a selfie, group photo, or snapshot into a natural portrait of the same people in the same clothes and place.
AI product photo generator: turn one product photo into an ad-ready image, or a white-background product photo when you ask for one.
AI image variations: get the same photo with one change inside it.
AI camera control: get the same scene from a new camera angle.
AI image to video: animate one image as the opening frame of a video.
AI character generator: create a reusable Character. See Create and Use Characters.
Each tool creates one result by default. If the direction explicitly asks for a number from 2 to 10, the results appear in one carousel and each image is charged at the normal per-image price.
Image tools default to a 4:5 result, and attaching a photo no longer changes the aspect ratio. Choose another model, aspect ratio, or resolution under Settings before creating.
Edit prompt restores only your optional direction. If you did not enter a direction, Edit prompt remains blank; private tool instructions are never shown.
If you are signed out, you can prepare a guided-tool draft before logging in. The photo stays private in that browser for up to 24 hours, and Typeshutter still requires a final Create click after login before uploading the photo or spending credits.
Notes
For best results, describe what should be visible in the image instead of only naming a style.
Create from image accepts one helper image at a time. HEIC and HEIF images are converted before prompt creation when supported.
While Typeshutter is creating the prompt, the prompt field shows Creating prompt... and is locked until the generated prompt is ready or the request fails.
For people or fashion images, generated prompts use broad, non-identifying details such as face angle, hair, expression, lighting, pose, and photographic realism. They should not ask Typeshutter to copy a specific person's identity or exact likeness.
Use reference images when the selected model supports them and you want the output to follow an existing visual direction. Typeshutter sends the complete reference photo for image generation, even when you choose a different output shape.
You can cancel a generation only while it is queued. Once its card shows Processing, the prompt box clears and the generation continues from the card.
[Screenshot placeholder: Create image form showing Describe your shot or create from image.]
[Screenshot placeholder: Creating prompt... loading state after choosing one helper image.]
Troubleshooting
If create from image does not appear, confirm the prompt field is empty and the page is not currently creating or polling a generation.
If Creating prompt... stays visible for too long, wait a moment and then retry with the image again. If the image format is unsupported, choose a PNG, JPG, WebP, HEIC, or HEIF image when supported by your browser.
If the generated prompt appears but Create remains disabled, confirm the prompt field is not empty and no reference upload error remains.
If people prompts produce repeated or overly polished faces, edit the generated prompt with broad realism cues such as face angle, hairstyle, expression, lighting, lens feel, and natural skin texture. Avoid wording that requests a specific person's identity, exact facial features, or detailed likeness.
If a prompt that asks for several images creates only one, add the number of images to the prompt.
If the result does not match your idea, revise the prompt with more concrete visual details.
If the generation fails, check your credits and try again with fewer references or another model.
If your prompt or the cancel button still appears while a generation card says Processing, refresh the page. If it keeps happening, contact support with the media ID, your browser, and the approximate time it happened.