# Make and prepare pictures in Image Maker

Generate or edit a picture, change it with the Remake tools, turn a character into a four-view sheet and send the result to 3D, step by step in Image Maker.

> Canonical: https://ai3d.studio/docs/image-maker

A 3D model is only as good as the picture it starts from. The [AI Image Maker](https://ai3d.studio/image-maker) makes that picture from a prompt or from your own images, cleans it up, and hands it to a 3D page with the picture already loaded. The picture and the 3D model are separate runs, and each is priced before it starts.

## Step 1: Choose a category and a model

The rail at the top of Image Maker switches between Create, Remake, Looks, Change and Helpers.

- Create starts from a text prompt. Every card shows the model's default price.
- Remake holds ready-made changes aimed at 3D, each on its own page. Step 4 lists them.
- Looks holds single looks and figure styles from those tools, such as Marble bust, Building-block figure, Anime and Claymation. A look card opens its tool already set to that look.
- Change starts from your own reference images plus an instruction. Each card says how many reference images the model takes.
- Helpers holds the single-purpose tools. Step 5 lists them.

The studio offers 19 image models. Each one can generate from text, and each but FLUX.2 Max can also edit from reference images. Nano Banana 2 carries the Default badge. Picking a card loads that model into the form below it.

![The Create rail of Image Maker: every model card shows its starting price, and Nano Banana 2 carries the Default badge](https://cdn.ai3d.studio/ai3d/image/docs/image-maker/step-1-v3.webp)

## Step 2: Describe one subject

1. Name the object and its material in the prompt, then ask for the whole subject in frame on a plain background with soft light. That is the picture an image-to-3D run reconstructs most cleanly.
2. Read the credit cost beside the Generate button and press Generate. Asking for several images runs one job per image at the card's price, and a failed job is refunded.
3. Avoid lettering on the object. Text drawn by an image model often comes out garbled, and it does not survive the 3D step.

![A prompt for one ceramic owl figurine on a plain background, the aspect ratio and resolution options, and the credit cost beside Generate](https://cdn.ai3d.studio/ai3d/image/docs/image-maker/step-2-v3.webp)

## Step 3: Send the picture on

A finished picture carries up to six buttons: [Turn into 3D](https://ai3d.studio/image-to-3d), [Remove background](https://ai3d.studio/background-remover), [Fix the lighting](https://ai3d.studio/remove-shadows-from-image), [Upscale](https://ai3d.studio/image-upscaler), [Make a character sheet](https://ai3d.studio/character-sheet-generator), which draws four views of a character for multi-view to 3D, and [Put in an A-pose for rigging](https://ai3d.studio/ai-pose-changer?preset=pose-a-pose), which redraws a character with its arms down at an angle, the pose auto-rig fits best. A picture never carries the step that made it: a cut-out has no Remove background, a relit picture no Fix the lighting and an upscaled one no Upscale. Turn into 3D opens Image to 3D with the picture loaded; nothing is charged for the 3D model until you run it there. [How to make a 3D model from a picture](https://ai3d.studio/docs/image-to-3d) continues from that page.

![A finished picture of a white sneaker with its next-step buttons: Turn into 3D, Remove background, Fix the lighting, Upscale and Make a character sheet](https://cdn.ai3d.studio/ai3d/image/docs/image-maker/step-3.webp)

## Step 4: Use the Remake tools

Remake changes one picture into something a 3D run wants. Each tool is its own page.

![The Remake rail: Photo to figurine, Pose changer, Character sheet, Style transfer, Part extractor, Consistent character, Sketch to image and Picture for 3D, each shown with a sample](https://cdn.ai3d.studio/ai3d/image/docs/image-maker/step-4-v2.webp)

| Tool                                                    | What it does                                                                     |
| ------------------------------------------------------- | -------------------------------------------------------------------------------- |
| [Photo to figurine](https://ai3d.studio/3d-figurine-maker)                 | A photo becomes a collectible figurine on a base.                                |
| [Pose changer](https://ai3d.studio/ai-pose-changer)                        | Redraws a character in a chosen pose, including T-pose and A-pose for rigging.   |
| [Character sheet](https://ai3d.studio/character-sheet-generator)           | One character drawn from four views, for multi-view 3D.                          |
| [Style transfer](https://ai3d.studio/ai-style-transfer)                    | Restyles a picture, for example as toon, anime, pixel art, claymation or bronze. |
| [Part extractor](https://ai3d.studio/image-part-extractor)                 | Pulls a helmet, a weapon or any other part out, redrawn on white or cut out.     |
| [Consistent character](https://ai3d.studio/consistent-character-generator) | The same character in new shots or in a scene you describe.                      |
| [Sketch to image](https://ai3d.studio/sketch-to-image)                     | Renders a rough drawing as a finished picture.                                   |
| [Picture for 3D](https://ai3d.studio/ai-image-for-3d)                      | A clean, single-subject picture made from a description.                         |

In the pose changer, style transfer and part extractor, put your own picture in the Reference Image field first; the pose, style or parts you pick come second. In the part extractor, Analyze this picture (1 credit) lists up to 12 parts for you to rename, untick or add to before anything is extracted. Each part is either redrawn whole (a Nano Banana 2 edit, 10 credits at the default size) or cut out of your picture's own pixels onto a transparent background (1 credit, and it cannot fill in what is hidden); you choose per part. Make 3D for all parts then sends every extracted picture to Image to 3D as one batch, on a plan that includes batch. The pose changer groups its poses as For rigging, Everyday and Action; a T-pose or A-pose result is the input auto-rigging expects, as [How to rig and animate a 3D character](https://ai3d.studio/docs/rig-and-animate) explains.

### Looks by id, from the API and MCP

Every figurine template, style look and pose is a preset with an id, such as `figurine-bust`, `style-anime` or `pose-t-pose`. The studio sends that id and the server writes the full prompt, so the text under What this sends, followed by your own instructions, is exactly what runs, and a preset costs what the model it runs on costs. An API or MCP call names a preset in `options.preset` with the picture in `options.image`, and may add `preset_addons`, `preset_background` and a short instruction in `prompt` of at most 500 characters; an instruction over the limit is refused rather than cut. Three presets have no page of their own: Free-form remake, Parts layout and Tabletop hero. Their cards under Looks run them in Image Maker's own form, with their extra touches above it: Parts layout can name each part, number the parts or lay them out as an exploded view, and Free-form remake asks what the picture should become. The [developer documentation](https://ai3d.studio/developers) lists the presets your key can use.

## Step 5: Use the Helpers

| Tool                                            | What it does                                               |
| ----------------------------------------------- | ---------------------------------------------------------- |
| [Background remover](https://ai3d.studio/background-remover)       | Cuts the subject out onto a transparent background.        |
| [Shadow remover](https://ai3d.studio/remove-shadows-from-image)    | Replaces harsh shadows and hot spots with even light.      |
| [Image upscaler](https://ai3d.studio/image-upscaler)               | Makes the picture larger with Topaz or with Recraft Crisp. |
| [Image to prompt](https://ai3d.studio/image-to-prompt)             | Writes a prompt that describes the picture.                |
| [AI material generator](https://ai3d.studio/ai-material-generator) | Makes seamless PBR maps from a description or a photo.     |

The [Prompt Builder](https://ai3d.studio/prompt-builder) is a chat that writes the prompt for you. Describe an idea in your own words, attach up to 3 reference pictures (JPEG, PNG or WebP, at most 4 MB each; a larger picture is shrunk in your browser first), and choose Quick, Standard or Careful. It asks at most two short questions, then writes one prompt with up to 3 changes you can try next. Use prompt fills the Image Maker or Text to 3D form, and Save keeps the prompt under My prompts on the [Prompt Ideas](https://ai3d.studio/prompts) page. The same chat opens from the Prompt Builder button under the prompt of any Image Maker or Text to 3D form. Replies are free and limited per model each day: 40 on Quick, 20 on Standard and 5 on Careful. Sending the same message again after a failed reply does not use up another one.

## From a character sheet to a multi-view model

A single picture hides the back of a character. The character sheet draws four views so a multi-view run can see them all.

1. Open [Character sheet](https://ai3d.studio/character-sheet-generator) and choose a pose: A-pose, T-pose (for rigging) or Keep the pose.
2. Add one picture of the character, whole body, on a plain background, and generate. One sheet is one image run that draws all four views. Splitting the sheet into views is free and happens in your browser.
3. Check the four views: Front, Left, Back and Right. If two views are swapped, change Image Makerel under one. If one view is wrong, use Redraw on that view alone.
4. Choose the Multi-view 3D model and press Make 3D model. [Multi-view to 3D](https://ai3d.studio/multiview-to-3d) opens with the views filled in, and nothing is charged for the 3D model until you confirm there.

## Turn a photo into a figurine

The [3D figurine maker](https://ai3d.studio/3d-figurine-maker) chains two runs. Pick a template, add a photo and restyle it, check the styled image, then press Make 3D. It shows the cost of both runs before you pay for either, and [Prepare a 3D model for printing](https://ai3d.studio/docs/3d-printing) follows the figure through to a print file.