WunderPaint documentation

Introduction #

What WunderPaint is, and who it is for.

WunderPaint turns the WordPress media library into a real design tool. Instead of only cropping and rotating, you get a complete layer-based editor: paint and shapes, rich text and typography effects, non-destructive filters and adjustments, masks and blend modes, vector and PSD support, AI generation and editing, dynamic templates, and an animation timeline, all in a full-screen editor that opens on top of wp-admin.

• Free core: the entire editor. Layers, tools, filters, effects, text, shapes, AI, export. No crippleware, no nag screens.
• Pro add-on: automation and scale. Dynamic templates run across your posts and products, batch image processing, series-from-CSV, the AI Content Generator, multiple brand kits and off-site backups.
• Extensions: optional studios that plug new generators and tools into the editor (3D text, 3D mockups, cinematic effects, patterns, maps, and more).

Opening the editor #

Several entry points, all opening the same full editor.

• Media library: hover or open any image, then click Edit Image.
• Media list / grid: the Edit Image row action and the attachment details button.
• “+ New” admin bar, then Image, and Media, then New Image for a blank canvas.
• Inside content: the Gutenberg image block, Elementor image controls, and any page builder that uses the WordPress media picker (see Edit in place).

Creating a new image #

Start from a blank canvas via Media, then New Image or the admin bar’s + New, then Image. Pick a preset size (social formats, print, custom), a background, or generate a starting image with AI. Everything you add lands on layers you can rearrange, mask and restyle at any time.

Edit in place New #

Edit any image without leaving the page you are building.

The full editor opens as a seamless overlay right where you are working. When you are done, the edited image drops straight back into the element you started from, as a fresh, bound copy, so the original media stays untouched everywhere else on your site.

1. Click Edit Image on the element (image block, Elementor image control, or the media picker).
2. The editor opens in an overlay on that exact image.
3. Edit, then click Apply (the button names where it returns to, for example Apply to Elementor).
4. The overlay closes and the element shows your edit. Close (next to Apply) or Esc cancels.

Interface tour #

Where everything lives.

• Toolbar (left): 16 tools with keyboard shortcuts. The options bar above the canvas adapts to the active tool.
• Canvas (center): your document, with rulers, guides, a configurable grid, snapping and a navigator. The eye in the status bar reveals whatever is parked outside the document frame, turning the space around the canvas into scratch space. Export still crops to the frame.
• Panels (right): Layers, Adjust, Effects, AI, History and the context-aware Properties panel.
• Menu bar (top): File, Edit, Image, Layer, Select, Filter, Text, View, Automation, Tools, Extensions and Help.
• Pages & tabs: multiple pages per document and multiple open documents.
• Fast access: a Quick Insert palette and a searchable Command Palette reach every asset and command by typing; savable Workspaces hide the controls you do not need.

Tools #

Sixteen tools, each with a keyboard shortcut.

Panels #

• Layers: reorder, group, mask, set blend mode and opacity, add layer styles and adjustment/fill layers.
• Adjust: live, non-destructive Brightness, Contrast, Exposure, Levels, Curves, Hue, Saturation, Vibrance, Temperature and more.
• Effects: filter presets, pixel effects, text styles and effects, all with single-glyph previews on adaptive backgrounds.
• AI: generate, edit, background removal, upscale, and quick actions.
• History: step back and forth through every change.
• Properties: context-aware controls for whatever is selected (position, size, text, fills, dynamic content).

Easy Mode #

An optional beginner shell (Help, then Easy Mode) that swaps the tool rail and panels for a simpler set while keeping the exact same documents. Good for occasional users; power users stay in the full editor.

Layers, masks & groups #

Non-destructive, Photoshop-style compositing.

• Layers and groups with 16 blend modes and per-layer opacity.
• Layer masks and clipping masks for non-destructive hiding and framing (drop an image onto a shape or text to fill it).
• Layer styles: strokes, shadows, glows and more, all editable later.
• Adjustment and fill layers that affect the layers beneath them.
• Layer comps: named snapshots of layout and visibility you can flip between.
• Smart Objects with non-destructive smart filters, plus true PSD import/export.

Adjustments & filters #

Adjustments are live and non-destructive: Brightness, Contrast, Exposure, Levels, Curves, Hue, Saturation, Vibrance, Temperature, plus shadow/highlight and white-balance controls.

Filters are one-click preset looks. You can also load .cube color LUTs for cinematic grading, and save your own looks to share.

Pixel effects & LUTs #

Over twenty pixel effects including Duotone, Vignette, Tilt-Shift, Halftone, Glitch, Glow, Scanlines, Grain and Color Noise, each with a live preview and adjustable settings. Combine them with LUTs and layer styles for a signature look.

Text & typography #

A real type engine, not just a text box.

• Rich text: mix sizes, fonts, weights and colors within one layer.
• Area and point text: click the Text tool to drop a growing box; paste a whole paragraph and it wraps and grows. Left, center, right and justify alignment.
• 50+ text effects with previews and settings: gradient and pattern fills, chrome, foil, 3D extruded perspective text, outlines, shadows, highlights and more.
• Curved and path text: bend text on an arc or flow it along any pen path.
• 70+ self-hosted fonts (no external CDN), plus your own uploaded fonts.
• Dynamic variables: mix {{tokens}} like {{post.title}} into text; they resolve per post when a template runs.

Shapes, pen & SVG #

• Shapes: rectangles, ellipses, polygons, lines, hearts, arrows and more, with fills, strokes, gradients and pattern fills.
• Pen tool: Bézier paths you can edit anchor by anchor.
• SVG: import SVGs as editable vector layers, and export your design as SVG.
• Vectorize: trace a raster image into editable vector shapes.

Selections #

Marquee, lasso and magic wand, plus Smart Select (a local AI subject / region selector). Feather, invert, grow, save and load selections; turn a selection into a mask, a new layer, or a crop.

AI features #

Cloud generation when you want it, on-device AI when you do not.

Cloud generation and editing

Create new images or edit existing ones with Google Gemini / Imagen and OpenAI. You enter your own API keys in Settings; prompts and (for edits) the image are sent to the provider you choose, only when you trigger an action. Keys stay on your server.

• Design Review: an art director's verdict on the design you have open, with the few changes that would make the biggest difference, answered in your editor language.
• Improve Text: stronger alternatives for the selected text layer, same language and roughly the same length. Pick one and it swaps in.
• Turn into editable vector: the model draws flat artwork and a local tracer turns it into real shape layers instead of pixels.
• 360° format: generates the spherical images the 360° Panorama tool opens.

Local, key-free AI

• Background removal and upscaling: run locally, no key.
• Blur Faces: find every face on a layer and pixelate it, locally, no key.
• One-click looks: Color Pop keeps the subject in color and drains the rest, Product Shot cuts out and drops the piece onto clean white with a contact shadow, Neon Rim glows along the edge in your brand colors, Speed Blur zooms the background past a sharp subject, and Depth Fog lays haze by real distance.
• Text Behind Subject: the magazine cover look, with an editable headline sliding in between background and subject.
• Smart Select, Depth Blur, local alt text, Split into Layers (OCR) and captions.

Local models download once to your own server (under uploads/wpie-models) and then run in the browser, so nothing is sent to a third party at run time.

Design Assistant

Describe what you want and the model art-directs three genuinely different layouts, rendered as fully editable layers you can then refine.

Charts, tables, QR & icons #

• Charts: bar, line, pie and donut charts from your own data.
• Tables: data, comparison and pricing tables.
• QR codes as editable layers, and 5,000+ Tabler icons.
• 1,900+ emoji as design elements.

One-click design tools #

Small, focused studios under Tools. Each solves one job and returns editable layers.

• 360° Panorama: describe a place and generate the whole sphere, or let a photo you already have become the entire scene around the camera. Look around it, place linked hotspots, and copy an embed that carries its own viewer, so the page it lands on needs no plugin.
• Collage & Photo Grid: a handful of photos becomes a clean grid, a justified mosaic, tilted polaroids, a filmstrip or a contact sheet, inserted as an editable group where every photo stays a normal layer.
• Color Schemer: build a five-color scheme with color theory, from an image or by chance, then push it site-wide or into a brand kit.
• Background Studio: generative full-document backgrounds with sliders and a seed, rendered locally.
• Screenshot Beautifier: wrap a screenshot in a rounded card with shadow, browser, window or phone chrome, and a generated backdrop.
• Mockup Generator: place your artwork on an ebook, box, DVD, report or monitor, rendered locally on a transparent background.
• Contrast Check: rate every visible text layer against the pixels actually behind it and flag anything below the WCAG recommendation.
• Vignette & Film: vignette, light leaks, film grain, dust, scratches and a retro date stamp on a single re-editable overlay layer.
• Batch Watermark: stamp a watermark onto many library images at once, with a live preview and saved settings.

Brand kits & assets #

Store your colors, fonts and logo in a Brand Kit and apply them across designs. Build a color palette from any image, keep a gradient preset library, save custom pattern tiles, and organize reusable assets. (The Pro add-on unlocks multiple brand kits.)

Media Library Manager Free #

A full-screen home for your whole media library: search by what images actually show, organize with folders and tags, fix metadata in bulk, and reclaim space.

Open it from the editor's Tools menu (Manage Media Library). It sits on top of the standard WordPress library and adds the tools a real image workflow needs, without sending anything to a third party unless you ask it to.

Find anything

• Semantic search: type what you are looking for (“red sneakers on concrete”, “team photo outdoors”) and the manager searches by what each image shows, not just its file name. This runs on a local image-search model you install once under Settings.
• Find similar: pick one image and surface near-matches across the library.
• Filter by media type (images or videos), orientation (portrait, landscape, square), size (small to huge), plus quick filters for missing alt text, unused and broken files.
• Sort by newest, oldest, title, or largest and smallest, in a cloud or list view.

Organize

• Folders and subfolders: group media into virtual folders by dragging. Folders are non-destructive, the underlying files never move, so nothing else on your site breaks. Suggest folders can propose a structure from visually similar groups.
• Tags: label images by hand, or let Auto-tag add tags with a local model.
• Smart folders: save any search or filter as a smart folder that keeps itself up to date.

Metadata, in bulk

Select one image or a hundred and let the Metadata Assistant generate titles, alt text, captions, tags and a mood, either on-device or through your cloud provider. You can also edit metadata by hand, auto-name files, and rename by pattern (with a counter and variables). Great for clearing a backlog of “(no alt text)” images.

Housekeeping

From the grid you can also set a featured image, insert an image into the open design, move a selection to the trash (restorable), or run a smarter upload that can describe images (alt, caption, tags) as they import.

The eye icon on a tile opens a large preview over the manager. Arrows fixed to the left and right edge of the window step through the current results, so you can click through a folder without moving the pointer, and the arrow keys do the same. From there you can insert the image into your design, copy its file URL, download the original, or jump straight into its detail view.

An image's detail view goes further: fix its metadata, replace the file while every link to it stays valid, or recrop its generated sizes by hand, per size, instead of living with the dead-center crop WordPress makes. Only the size files are rewritten, the original is never touched.

What is inside a file

The detail view's File data tab reads what the file itself carries, which is not the same as what the library shows you. A photo from a phone records the camera and lens, the exposure, the software that has touched it since, and usually the exact spot on earth it was taken at. None of that appears anywhere in WordPress, and the original file stays reachable under its own address once you publish it.

Coordinates are turned into a place name and a rough distance, worked out on your own server against the place index that ships with the plugin, so finding out where a photo was taken never contacts an outside service. Two buttons take things back out again: Remove location leaves the camera details alone, Remove all metadata clears every embedded block. Both run on the original and on every size generated from it, and neither re-saves the picture, so image quality is exactly as before. The previous files are kept as a version you can restore.

Templates & dynamic content #

Save a design as a reusable template, and turn text and image layers into dynamic placeholders with {{tokens}} and bindings: post title, excerpt, author, featured image, WooCommerce product name, price, image and more. The design becomes a reusable layout; the content fills in per post.

Saving & export #

• Save to Media Library as a new version of the original (with rollback).
• Save as New Image: a new attachment.
• Export: PNG, JPEG, WebP, PSD, plus animated GIF, APNG and WebM from the animation timeline, and SVG.
• Portable projects: download a .wpie project file.
• Export Carousel: slice a wide document into a numbered sequence of equal tiles, for a swipeable social carousel.
• EU AI label: place one of the European Commission's official emblems for AI disclosure into the picture, in the color, corner and size you pick, and optionally record the matching IPTC entry in the file as well.

Labelling AI involvement

The save and export dialog offers the three official emblems the European Commission publishes: a basic one, Fully AI-generated and Partially AI-modified, each in black, white and half-opacity versions. The list starts on None, and nothing is ever applied unless you pick a motif yourself. The editor does not inspect your design and never suggests one, because the disclosure rules are about content that could be taken for a real photograph, not about every illustration, and only you know which one you made.

A second box, empty unless you tick it, additionally writes the matching IPTC digital source type into the file. That is the vocabulary picture agencies and image search read, so a picture can declare its origin even where the emblem itself has been cropped away. The same emblems also sit in the Asset Library under Brand Kits, if you would rather place one as an ordinary layer and position it yourself.

Re-editable images #

Your layers come back, even after the image is inserted as a flat PNG.

When you save, the editor also stores the raw, editable layer project alongside the attachment (as long as “Keep editable project file” is on, which it is by default). The image that appears on your site is the flat render; the layers live with the attachment.

Click Edit Image on that image later, from the media library, Gutenberg, Elementor or any picker, and the editor restores your full layered version, exactly where you left off. This works for the bound copies created by edit-in-place too.

Guide: your brand kit, used everywhere Free #

Define your corporate identity once, then reach the same colors, fonts and logo in every design, in the AI, and in the extensions.

1. Open Brand Kits and build a kit: your brand colors (up to 12), your fonts (up to 3), your logo, and an optional company profile. In the dialog's own words, this is the source for quick picks, templates and AI content.
2. In any design, your kit appears as a captioned swatch group in every color picker, so you fill shapes, text and strokes straight from your palette instead of pasting hex codes.
3. Your brand fonts are offered in the text properties, and your logo is a reusable asset you can drop onto any layout.
4. The extensions read it too: Pattern Generator, Generative Studio and the 3D studios offer your kit as a palette, so generated art stays on brand.
5. The AI uses it as well, see the next guide.

Guide: CI-compliant AI, your style guide in every prompt Free #

Guide: dynamic templates with variables Free #

Design once, and let each post or product fill in its own text and images.

1. Design your layout: background, headline, and image placeholders, on layers as usual.
2. Put variables into text with the variable button (a small {…} in the text properties), or type tokens directly, for example {{post.title}}. Available variables include {{post.title}}, {{post.url}}, {{post.excerpt}}, {{post.date}}, {{post.category}}, {{post.tags}}, {{post.readingTime}}, {{author.name}}, and the WooCommerce fields {{wc.name}}, {{wc.price}}, {{wc.discount}}, {{wc.sku}}, {{wc.stock}}, {{wc.rating}}.
3. Bind an image layer to a source in the Properties panel under Dynamic content, for example the post featured image or {{wc.image}}. For bound text, set the Min/Max font-size so it auto-fits.
4. Check it against a real post: View, then Preview with Post.
5. Save it: File, then Save as Dynamic Template.

Guide: auto featured images from post content Pro #

Generate an on-brand featured image for a whole batch of posts.

1. Open Automation, then Featured Images.
2. Pick the posts (filter by category; the list shows how many already have a featured image).
3. Choose a Dynamic Template. You can also assign templates per category, with the top selection as the fallback, or pick No template for finishing only.
4. Optionally pick a Brand kit (its colors and fonts steer the AI background and fill brand variables) and an Action.
5. Turn on AI Background to generate a background from the post, and optionally Enhance design with AI and Generate SEO title and description.
6. Press Process. Each result is set as the post featured image, which is also used as the OG / Twitter image.

Guide: dynamic images on the front end Pro #

Put a design on the front end that fills with live data, without opening the editor.

Three modes, each available as a block and a shortcode. Pick by how the data changes:

Guide: self-updating shop and content graphics Pro Extension #

One setup, then the graphic maintains itself: new products, fresh reviews, the latest episodes.

1. Open Extensions, then Showcase Studio and build the graphic from live data: your products (with on-sale, in-stock or best-selling filters), posts, reviews, team, a feed or an API. Pick a built-in layout or repeat your own card via Custom Card.
2. Save the document as a template via File, then Save as Template.
3. Put it on the site with the Dynamic image block or [wpie_dynamic_image template="…"]. The baked image keeps a stable URL, so it can also be embedded in newsletters or external pages.
4. In Automation, then Featured Images, enable Auto-refresh dynamic images. A daily schedule marks the graphic stale, and the next wp-admin visit re-renders it quietly in the background. Shop events (price changes, sale start and end, new reviews) refresh it on their own.
5. Optional: a campaign window swaps the design for a date range (campaign="…" campaign-from="2026-11-24" campaign-until="2026-11-27"), and the webhook field pushes every fresh render to Zapier or Make.

Guide: WooCommerce product images Pro #

Finish a product gallery and add an AI infographic, in a batch.

1. In Automation, then Featured Images, select your products.
2. Choose a Gallery template and/or an Action plus Enhance. Every gallery image runs through the enhance/action finishing and the gallery template, and the product gallery is then replaced with the results.
3. Optionally turn on Append product infographic: one AI infographic built from the product name, price, short description and rating, added to the end of the gallery.

Guide: a series of images from CSV Pro #

A mail merge for graphics: one saved image per row.

1. Design a template with {{placeholders}} in the text layers.
2. Open Automation, then Generate from CSV and upload a CSV whose column names match your placeholders.
3. Generate. You get one saved image per row.

Great for certificates, badges, event line-ups and per-product cards.

Guide: 3D product mockups and 360° spin Extension #

Wrap your design onto a 3D product and export a turntable animation.

1. Open Extensions, then 3D Mockup Studio and pick a scene or model (mug, bottle, shipping box, book, phone case, t-shirt, poster and many more).
2. Apply artwork per surface: each surface can use its own design, a layer, the whole document, or a template.
3. Adjust the camera angle.
4. Insert it as an editable layer; reopen it anytime via Edit Mockup.
5. Export a 360° spin sheet for an animated turntable of the product.

Guide: a dynamic product template on a 3D mockup Pro Extensions #

One template that renders an on-brand, 3D product image for every product. This is the capstone: it chains the brand kit, dynamic bindings, the 3D mockup and dynamic 3D text.

1. Design a dynamic template. With your brand kit active, build a product layout and bind its layers to product data: {{wc.name}}, {{wc.price}}, and the product photo {{wc.image}}. Use your brand colors and fonts, then save it as a Dynamic Template.
2. Wrap it onto a product. Open 3D Mockup Studio, pick a product (a box, a bottle, a phone case, a book), and set the surface artwork to Template, choosing the template you just made. The studio lists it as Template: your name among the artwork sources.
3. Add dynamic 3D text. With 3D Text Studio, add an extruded 3D text layer whose content is a token, for example the price via {{wc.price}}. It is a generator layer, so it re-renders per product too.
4. Insert and save. Insert the mockup as an editable layer (reopen anytime via Edit Mockup), then save the whole design as a dynamic product template.
5. Run it per product. Through WooCommerce automation or Featured Images, each product renders its own shot: both the 3D text and the mockup's template artwork resolve against that product.

Guide: sprite sheets and 360° spins Extension #

Three studios can bake a turntable. This is what the result actually is, and what you do with it.

3D Mockup Studio, 3D Text Studio and 3D Gallery Studio all offer a 360° Spin card, and all three hand you the same three things. Which one you want depends on where the spin has to run, not on which studio you started in.

What a sprite sheet is

A sprite sheet is a flip book in a single file. The frames sit side by side in a grid, and a player shows one cell at a time and steps to the next one as the visitor drags. Twenty-four frames become five columns by five rows, with the last cell left empty. At the default frame size that is one image of 1920 by 1920 pixels.

Which of the three to take

1. Insert sprite sheet when you still want to work on the picture. It arrives as one layer, so you can crop it, put your brand on it and export a single PNG. Take this one if you already have your own player.
2. Viewer (HTML) when the spin simply has to work on a page. One self-contained file with the sheet embedded, so there is nothing else to upload and no request leaves the page. It turns by itself until the first drag, then follows the pointer. On a phone a horizontal drag turns it and a vertical swipe still scrolls the page.
3. Frames (ZIP) when a 360 player is already in place. One PNG per frame, named frame-01.png upwards and zero padded, in turning order, so a viewer that sorts by file name gets the rotation right with no renaming.

Playing a sheet yourself

Give an element the size of one cell, set the sheet as its background, and move the background position by one cell per frame. Frame i sits at -(i % cols) * cell across and -floor(i / cols) * cell down.

Both numbers matter: because the frames sit in a grid rather than one long strip, a plain CSS steps() animation is not enough on its own, since it walks a single axis. A grid needs a few lines of JavaScript, or a player that understands sheets. If all you want is drag to rotate, take the Viewer (HTML) download instead, which already does exactly this.

The per-control reference lives in the glossary: Sprite sheets: what to do with a 360 spin.

Guide: extruded 3D text and logos Extension #

Turn a headline or a logo into a real 3D object, lit and reflected, and drop it into your design as a transparent layer.

1. Open it. Extensions, then 3D Text Studio. It starts in text mode; switch to 3D Logo to extrude an SVG file instead, or a vector layer straight out of your document, multi-colour fills and all.
2. Type and shape. Enter the text, then set depth, bevel and curve. Twelve bundled display faces are there to start with, and any font already in your editor library works too, including your own uploads.
3. Pick a material. Twelve presets, from gold and chrome to glass, neon glow and hologram. Each one sets the surface plus its own face and side colours, which you can then override.
4. Style one letter at a time. Click a single letter in the preview and give it its own colour, material, size, depth, shift and rotation. This is what makes a word look designed rather than generated.
5. Light it and judge it in place. Set brightness, shape the cast shadow, add a mirror floor. Turn on Show document to see the text over your real artwork instead of a checkerboard.
6. Insert. It lands as a transparent PNG layer, centred and scaled to fit, and stays editable through Edit 3D Text.

Guide: photo galleries as 3D walls Extension #

Arrange a set of photos as a designed wall, tilt it in real 3D, and take it out as a picture, a video loop or something the visitor can drag.

1. Choose the photos. Either a fixed set you pick by hand, or a dynamic set: a post's featured image and all its images, a WooCommerce product with its gallery, or simply the newest posts, products or uploads.
2. Pick a layout. Grids, justified rows, masonry, mosaic, honeycomb, circles, hearts, stars, coverflow, a curved wall, a carousel ring, a card deck or scattered polaroids. Each one adapts to how many images you gave it.
3. Style the tiles. Spacing, rounded corners, frames from hairline to polaroid, drop shadows, and optional captions on pills.
4. Tilt it. Orbit the whole wall in real 3D, bend a flat wall into depth, set the camera and choose what sits behind it. Twenty-four presets are there if you would rather start from a finished look.
5. Give it motion, if you want one. Orbit, spin, sway, float, pulse, drift or a cinema pan. The motion is what the video and embed exports record.
6. Take it out. A baked layer, a seamless video loop as MP4 with a WebM fallback, an interactive drag-to-rotate HTML embed, or a 360 spin. Reopen any of it through Edit 3D Gallery.

Guide: ready-made 3D objects Extension #

Drop a real 3D object into a design, repaint it, light it, and bake it as a transparent layer. Bring your own models, or have one built.

1. Pick an object. Over 150 in nine categories: Shapes, Symbols, Party, Commerce, Food, Tech, Seasonal, Nature and Abstract, each one a live thumbnail rather than a flat icon.
2. Or bring your own. Import a textured GLB (binary glTF 2.0 with embedded textures). It is parsed locally in your browser, and it goes into a server-side library that the whole site shares.
3. Or have one made. With a Meshy key, any media-library image or a text description becomes a textured 3D model, saved straight into your library.
4. Repaint it. Fourteen material presets from gold and chrome to glass, neon glow and toon, or keep the object's own colours.
5. Aim and light. Drag to orbit, wheel to zoom, one-click camera angles and light moods, a soft ground shadow and an optional floor reflection. Show document puts your real artwork behind it so you judge it in context.
6. Insert. It bakes as a transparent PNG layer, centred, and stays editable through Edit 3D Object: object, material, camera and light all come back.

Guide: books, magazines and paper that turns Extension #

Everything made of paper that turns, curls and peels, as a real 3D object you can take hold of.

1. Pick the model. Twenty-four kinds of paper thing: a hardcover book you can actually leaf through, a limp glossy magazine, folded maps, menu cards, jigsaw puzzles, envelopes, and a strewn pile of wavy sheets.
2. Feed it pages. Straight from the media library, or from a PDF that is read page by page inside your browser and never uploaded anywhere.
3. Shape the paper. Page shape, boards and backing, wave and corner-curl sliders, plus peel and pile controls for the loose models.
4. Turn it by hand. Grab a page by its corner and drag: that page turns. Drag away from a corner and the whole object turns instead. The wheel moves the camera closer.
5. Set the scene. Light, shadow, camera and background, so the object sits in your design rather than floating on a checkerboard.
6. Take it out, two ways. Freeze the current state as a picture with a transparent background, so a curled sheet or a half-peeled sticker can sit inside any design. Or copy one block of HTML that carries its own viewer and runs on any website.

Pro overview Pro #

The Pro add-on is about automation and scale.

WunderPaint Pro is a separate plugin that installs alongside the free editor. The whole editor stays free; Pro adds the features that turn it into an automated image pipeline:

Pro also ships dynamic image blocks and shortcodes ([wpie_dynamic_image], [wpie_live_badge], [wpie_personalized]) and WooCommerce helpers (gallery processing, product infographics).

Automation Pro #

Point a dynamic template at a set of posts or products and generate an on-brand image for each, with the right title, price or featured image filled in. Featured Images can be generated automatically from a post’s content, so a whole archive gets consistent, CI-correct social imagery.

Image Processor (batch) Pro #

Select images from the library (or paste IDs), choose a format (WebP, Web JPEG, lossless PNG), quality, resize and optional watermark, or run a recorded Action, then process them all in one pass, overwriting in place (versioned) or as new attachments. Quick presets like WebP Small make it a practical bulk optimizer.

Series from CSV Pro #

A mail merge for graphics: put {{placeholders}} in your text layers, upload a CSV, and get one saved image per row. Perfect for certificates, badges, per-product cards and event line-ups.

AI Content Generator Pro #

Generate structured, SEO/GEO-minded articles and content in your chosen output language, with brand-aware imagery to match, from inside the same tool you design in.

Free vs Pro #

Extensions overview Add-on #

Optional studios that plug new generators and tools into the editor.

Extensions are one-time add-ons that appear inside the editor (Extensions menu) and insert their output as normal, fully editable layers. They use the same public Extension API any developer can build on (see the developer docs).

3D studios Extension #

Motion & generative Extension #

Patterns & design Extension #

Maps, charts & more Extension #

Print & craft Extension #

Documentation & type Extension #

Settings #

Under Settings, then WunderPaint: default provider and API keys, local AI model downloads, theme, default canvas size, versioning (how many versions to keep), watermark, fonts (self-hosted plus optional Google fonts), stock providers, and the editor capability (who may use it).

Help lives inside the editor as well: the searchable Handbook, a Help Assistant that answers questions and builds small guided tours, and Help, then System Status for exactly what your device and server offer.

Keyboard shortcuts #

Every tool has a single-key shortcut (V move, B brush, T text, P pen). Common actions: ⌘S save, ⇧⌘E export, ⌘Z undo, Esc cancel or close. Shortcuts are remappable in Preferences.

FAQ #

Do I need AI API keys?

No. The whole editor and the local AI features (background removal, upscaling, smart select) work without any key. Keys are only needed for cloud generation with Gemini/Imagen or OpenAI.

Where are my API keys stored?

On your own server. They are never exposed to the browser, and calls to a provider happen server-side, only when you use that feature.

Does it change my original files?

No. Your original stays untouched until you choose to save. You can save as a new version (with rollback) or as a new attachment.

Can I re-edit an image later with its layers?

Yes, see Re-editable images. The editable project travels with the attachment.

Which page builders are supported for edit-in-place?

Elementor and the Gutenberg image block natively, plus any builder that uses the WordPress media picker (Bricks, Divi, Beaver, and the classic editor).

Privacy & data #

Nothing leaves your site on its own.

WunderPaint has no analytics, no tracking and no telemetry. It does not phone home, it does not register your site anywhere, and it never sends your images or your content to us. Editing, compression, search, the Media Library tools and the local AI helpers all run on your own server and inside your own browser. The bundled font families are self-hosted, so the editor contacts no font CDN by default.

Everything that can reach an outside address belongs to a feature you use on purpose, and most of those stay completely silent until you enter an API key of your own. Every one of them is listed under External services below, with the exact data it transmits and when.

Where things are stored

• Images and projects stay in your Media Library on your own server. The original file is untouched until you choose to save.
• API keys live in your WordPress options, are never handed to the browser, and are only used server-side for the feature you trigger.
• Local AI models are downloaded once to uploads/wpie-models and served from your own site afterwards. They then run in the browser, so the images they process never leave the machine you are working on.
• Caches for map geometry, place lookups, feeds and the extension catalog sit on your own server and hold only the answer, never anything about the person who asked.

External services #

This is every outside address any part of WunderPaint can contact. Each belongs to a feature and is only contacted when you use that feature, and a service that needs an API key stays silent until you enter that key under Settings, WunderPaint, AI Providers. Unless an entry says otherwise the request is made server-side by your WordPress site, so the service sees your server's IP address and not your visitors'.

WunderPaint extension catalog

Host delivery.wp-image-editor.com. When an editor user opens the Extensions manager and the one-hour cache on your site is cold. Sent a plain GET request for the file catalog.json. No API key, no site data, no user data, no identifier of any kind. Received the list of available extensions with names, versions and descriptions, cached on your site for one hour. The free plugin never downloads or runs extension code by itself: the manager shows a download link, your own browser fetches that ZIP from the same host, and you install it manually.
Privacy policy · Terms

Stock photo services

Pexels, Pixabay and Unsplash all work the same way. Nothing is sent until you enter that service's own key, and then only when you search under Assets, Stock Images. Your server sends the search text you typed, the page number and the number of results per page, together with your key. The result thumbnails are then shown straight from the service's image servers, so the browser of the person searching contacts them directly and its IP address is visible there. When you place one of the pictures in your design, your server downloads that file into your own Media Library.

• Pexels (api.pexels.com, images.pexels.com). The key travels in the request header. Privacy policy · Terms
• Pixabay (pixabay.com, including its image servers). Also sent: the image type you picked (photo, illustration or vector) and a safe-search flag. The Pixabay API accepts its key only as a query parameter, so unlike the other two the key travels in the request URL. Privacy policy · Terms
• Unsplash (api.unsplash.com, images.unsplash.com). Also sent: a content filter set to high. The access key travels in the request header. Privacy policy · Terms

Cloud AI providers

Google Gemini, OpenAI and Anthropic Claude all work the same way. Nothing is sent until you enter that provider's API key yourself, and then only when you trigger the action. Your key travels with every such request. What leaves your site depends on the action and never on anything else:

• Image generation sends your prompt.
• Image editing sends your prompt and the current image; inpainting and outpainting additionally the mask you painted; variations the source image alone.
• Alt text and image descriptions send the image and the language you want.
• Design Assistant, design review, improve text, gradient, lockup and vector suggestions send the brief or wording you wrote, the canvas size, and, if you filled one in, your brand kit of colors, font names, company name, industry, tone of voice and company description.
• SEO suggestions send the post title and excerpt.
• Text lockups send the wording of the text layer, its style note and its box size.
• An extension using the generic text endpoint sends the prompt it built.
• The connection test in Settings sends your key alone to the provider's model list.

• Google Gemini (generativelanguage.googleapis.com). Runs image generation and editing, inpainting and outpainting, 360° panorama generation, alt text and image descriptions, and all the text actions above. Privacy policy · Terms
• OpenAI (api.openai.com). Runs image generation and editing, inpainting and outpainting with a true mask, variations, alt text and image descriptions, and all the text actions above. Privacy policy · Terms
• Anthropic Claude (api.anthropic.com). Runs alt text and image descriptions, and all the text actions above. Claude is never used to generate or edit images here, so no image is sent to it except the one you ask it to describe. Privacy policy · Terms

OpenStreetMap Nominatim

Host nominatim.openstreetmap.org. When only when an editor user searches for a place inside one of the map extensions (Map Posters, Star Map, Route Visualizer) and the answer is not already cached on your site. Sent server-side: the place name you typed, the number of results wanted, the language for the answer, and a user agent line that identifies the plugin with its version and contains your site address, because the Nominatim usage policy requires requests to identify themselves. No API key, no account, no personal data. Results are cached on your own server.
Privacy policy · Usage policy

OpenStreetMap Overpass API

Hosts overpass-api.de, plus the mirrors overpass.kumi.systems, overpass.private.coffee and overpass.osm.jp, which are tried one after another if the first does not answer. When only when a map extension actually draws a map and the geometry is not already cached on your site. Sent server-side: a query containing the coordinates of the map section you are drawing, the requested level of detail, and the same identifying user agent with your site address. No API key, no account, no personal data. The street, water and park geometry that comes back is cached on your own server. Map data © OpenStreetMap contributors, ODbL.
Privacy policy · Terms

Meshy

Host api.meshy.ai, and the download addresses it returns. When only if you have entered a Meshy API key and you generate a 3D model in the 3D Objects studio. Sent server-side: for text to 3D the text prompt you typed together with the generation settings, for image to 3D the image file you selected, and in both cases your Meshy API key. Your site then asks Meshy repeatedly whether the job is finished, and when it is, downloads the finished model file and its preview picture into your own site. The connection test in Settings asks for your account balance and sends nothing but your key.
Privacy policy · Terms

Hugging Face

Host huggingface.co. When only when an administrator presses the download button for a local AI model under Settings, WunderPaint, Local AI Models. Sent a request for the file list of that model repository and then a request per model file. No API key, no account, no site data, no personal data. This is a one-time server-to-server download; afterwards the model files are served from your own site, and the in-browser AI runtime is configured never to fetch models remotely, so the browsers of the people using the editor do not contact Hugging Face at all.
Privacy policy · Terms

Google Fonts

Hosts fonts.googleapis.com, fonts.gstatic.com. The plugin ships ten font families with it and contacts no font CDN by default. Google is contacted only in the two cases you choose yourself:

• An administrator downloads additional families under Settings, WunderPaint, Fonts. Your server asks fonts.googleapis.com for the stylesheet of the family and requested weights, then downloads the matching woff2 files from fonts.gstatic.com into your uploads folder. Sent: the family name and weights, nothing else. This happens once, server to server, and afterwards those fonts are served from your own site.
• An administrator switches on the Google Fonts CDN option. The editor then loads font stylesheets from fonts.googleapis.com in the browser while people work, which makes their IP address visible to Google.

Exported images are unaffected either way, because text is rendered into pixels on your side.
Privacy policy · What Google Fonts logs

Feeds and data sources you enter yourself

Host any address you type. When only when you point a dynamic layer or a repeater at an RSS or Atom feed or at a JSON endpoint, for example a podcast feed, a YouTube channel feed or the now-playing endpoint of a web radio. Sent server-side: a plain request for exactly the address you entered, so that host sees your server's IP address and nothing else from your site. The answer is analysed on your server and cached there briefly. No data of yours is transmitted, and no such request happens unless you enter an address.

Media on external storage

Not a third-party service, but worth naming: if your Media Library is offloaded to external storage, opening such an image in the editor makes your site fetch that file from your own storage address server-side. This only ever concerns attachments of your own site that the current user is allowed to edit.

Freemius Pro

Host api.freemius.com. When only in WunderPaint Pro, which uses Freemius as its license and payment provider: when you activate or deactivate your license, when WordPress checks for a Pro update, and when it downloads that update. Sent your license key, your site address and the plugin and WordPress versions. Freemius also offers an opt-in screen on first activation that would additionally share your administrator email address and basic environment data; it tells you exactly what it sends before you decide, and you can skip it and keep using Pro. The free plugin does not contain Freemius at all.
Privacy policy · Terms

Premium extension downloads Pro

Host delivery.wp-image-editor.com. When only in WunderPaint Pro, when you install a premium extension from the Extensions manager. Sent server-side: the slug of the extension you picked, your license key and your site address, so the server can check that the license covers the download. Received the extension ZIP, which your site installs. Nothing about your content, your images or your users is transmitted.
Privacy policy · Terms

The local AI runtime

The runtime behind the local AI features (ONNX WebAssembly and transformers.js) belongs to the plugin instead of being pulled from a CDN, so Background Removal, Smart Select, Depth Blur and local alt text run entirely in your browser and your images never leave it. The only external touch is the optional one-time model download described above.