Viddo AI Image Canvas V1: Click to Edit, Doodle to Create

Stop Writing Prompts Like Code

If you've used any AI image editor in the last two years, you know the drill.

You open the tool. You start typing: "remove the trash can on the left, replace it with a small potted plant, keep the lighting natural, match the original style — please don't change anything else." You hit generate. You get something almost right. You tweak the prompt. You try again. You give up and open Photoshop anyway.

The promise of AI photo editing was that anyone could do it. The reality is that writing good prompts feels like writing code. You spend more time *describing* the edit than *making* the edit.

That's the problem Viddo AI Image Canvas V1 is built to solve.

Image Canvas V1 doesn't eliminate prompts. It removes the part that hurts: visual description. Instead of *typing* "the gray sofa on the left," you *paint* the sofa and the editor names it @Mark01. Instead of *typing* "the green velvet chair from the reference photo," you upload the photo and reference it as @image01. The prompt that remains is short, declarative, and English-like:

> *"Replace the sofa in @Mark01 with @image01."*

That's it. Seven words. The rest of the work is done by the model understanding the visual tags

 

Why We Built Image Canvas V1

Three problems kept showing up in creator conversations. We heard them from ecommerce sellers, social media managers, designers, and small business owners.

The Prompt-Length Problem

Every "easy" AI photo editor asks for a prompt. Detailed prompts. Long prompts. Prompts that have to mention style, lighting, perspective, color palette, and what *not* to change. For non-writers, this is a wall. For anyone who's ever tried to describe a sofa in 50 words, it's exhausting.

The irony: the AI already understands what an image *is*. The bottleneck isn't the model — it's the text interface.

The Precision Problem

Most AI image editors work on the *whole image*. You can't say "change only the shirt" without accidentally changing the background. The result is regenerations that look totally different from your original — which is fine if you're brainstorming, useless if you're a seller trying to keep your product consistent.

Image Canvas V1's Mark mode solves this with targeted inpainting. Paint directly on the area you want to edit. The model edits *only* that area. Everything else stays pixel-identical.

The Doodle Gap

Sketch-to-image tools have been around for a while, but they usually treat sketches as loose references, not as blueprints. You doodle a chair. The AI gives you a beautiful chair that has nothing to do with what you drew.

Image Canvas V1's Doodle mode is different. Doodle *on top of* the image you want to change. Sketch the new shape. The model fills it in with photorealistic detail. Your rough square becomes a perfect window. Your wobbly line becomes a believable lamp.

This is what "doodle to image AI" should have been all along.

 

What's New in Image Canvas V1

Image Canvas V1 is a complete rethink of how AI photo editing should work in a browser tab.

Two Modes: Mark and Doodle

The canvas has two distinct modes, and you switch between them with one click.

Mark mode — Precision editing for local changes.

· Pick a color (red, green, blue, purple, or white).

· Paint directly on the area you want to change.

· The paint automatically becomes a named tag: @Mark01, @Mark02, etc.

· Reference the tag in your prompt: *"replace @Mark01 with a leather armchair."*

Doodle mode — Free-form sketch-to-image.

· Doodle directly on the image: a new window, a missing object, a fresh background.

· The model interprets your doodle as the *target shape* and fills it in with photorealistic detail.

· Works especially well for "I want to add this thing here" use cases.

You can mix Mark and Doodle in the same edit. Mark the area you want to change, doodle what's coming next, and generate.

The @ Tag System (The Secret Sauce)

Every reference inside Image Canvas V1 is a tag. You reference tags in your prompt like you'd reference variables in code.

The three tag types:

Tag Type	What It Does	Example
`@Mark`	References a marked region on the image	`replace the floor in @Mark01 with @image02`
`@image`	References an uploaded reference photo	`paint the wall in @Mark02 to match @image01`
`@Hex`	References a color swatch	`change the door color in @Mark03 to @Hex01`

 

You don't have to remember tag names. Type @ and the editor lets you reference any of your marks, uploaded images, or color cards. Pick what you want. Drop it into your prompt.

This is what "tag-driven prompt editing" actually means in practice. You *do* write a short sentence, but most of the descriptive work is replaced by visual tags. Most Image Canvas V1 edits use prompts of 5-15 words.

The Prompt Editor accepts the same English-like sentences you'd write in any other tool. The difference is that @Mark, @image, and @Hex carry the visual context the model needs. The placeholder inside the editor reads:

> *"Type @ to insert tags, e.g. replace the sofa in @Mark01 with @image01…"*

That's a real prompt you can paste into the editor. It works.

Precision Controls

For when you want more control:

· Aspect ratio: 1:1, 3:4, 4:3, 16:9, or Custom

· Resolution: 1K (default), 2K, or 4K

· Guidance strength: a slider (default 0.75) that controls how strictly the model follows your tags. Higher = more faithful to your doodles. Lower = more creative liberty.

If your doodle isn't quite right and the AI is over-interpreting, slide guidance *up*. If the AI is being too literal and you want more natural variation, slide it *down*.

 

How to Use Image Canvas V1: A 3-Step Tutorial

You can do your first edit in under two minutes.

Step 1 — Upload and Mark Your Edits

1. Open [viddo.ai/image-canvas](https://viddo.ai/image-canvas).

2. Click "Upload an image" and pick any JPG, PNG, or WebP file.

3. Select a color from the Mark Color palette (red is the default).

4. Paint over the area you want to change. A red mask appears. The right sidebar shows your mask as @Mark01.

Want multiple regions? Just pick a different color and paint again. Each region becomes its own tag.

Step 2 — Add @ Tags and a Short Prompt

1. In the right sidebar, find the Prompt Editor — it's the box under Color Cards.

2. (Optional) Upload reference images via "Reference Images". Each becomes @image01, @image02, etc.

3. (Optional) Add custom color swatches via "Color Cards". Each becomes @Hex01, @Hex02, etc.

4. Type @ anywhere in the prompt to bring up your available tags.

5. Write your edit instruction in plain English. Example:

> *"Replace the gray couch in @Mark01 with the green velvet chair from @image01."*

That's a 14-word prompt that would take 50+ words in a traditional editor.

Step 3 — Generate, Compare, Export

1. Pick your aspect ratio and resolution.

2. (Optional) Adjust the Guidance Strength slider.

3. Click the big "Draw · Generate" button.

4. Watch your edit appear in the Latest tab. Switch to History to revisit previous generations.

5. If the result isn't quite right, click Undo or Redo, tweak your prompt, and regenerate.

6. Happy with the result? Hit Export to download your edited image.

That's the entire workflow. Three steps. No plugins. No tutorials.

 

4 Real Use Cases for Creators

Image Canvas V1 is built for the people who need to ship visuals every day — not for hobbyists playing with prompts. Here's where it shines.

1. For E-commerce Sellers

You're launching a new product. You have 40 product photos on white backgrounds. You want them on a marble countertop with soft natural lighting. You don't want to retake them.

With Image Canvas V1:

· Mark the white background as @Mark01.

· Upload a reference photo of the marble surface as @image01.

· Prompt: *"Replace the background in @Mark01 with the marble surface in @image01."*

· Generate. Get a believable product shot in 30 seconds.

Repeat for all 40 photos. Same prompt, same tag, same result. Consistency without a studio.

For color variations: paint just the product itself as @Mark01, reference a color card @Hex01, and prompt: *"Repaint the product in @Mark01 with the color in @Hex01."* Now you have a full color variant line from a single photo.

2. For Social Media Creators

You're a creator who shoots video and pulls frames for thumbnails, Reels, TikTok covers. Most frames have random objects in the background — your roommate's laundry, a parking meter, a stranger's elbow. Cropping helps. Erasing helps more.

With Image Canvas V1:

· Mark the unwanted object as @Mark01.

· Switch to Doodle mode and sketch what should be there (a clean studio backdrop, a plant, anything).

· Prompt: *"Replace @Mark01 with my doodle."*

· Generate. Done. No more cloning, no more lasso tool, no more "good enough."

For new product placements: doodle a coffee cup onto a flat-lay scene. The AI paints a photorealistic mug into your photo. Great for thumbnail A/B tests and styling mockups.

3. For Real Estate Agents

You've got a great listing photo — except the previous owners' ugly couch is in every room. Or the wall is a color you can barely photograph. Or the staging is from 2012.

With Image Canvas V1:

· Mark the old furniture as @Mark01.

· Upload a reference photo of a modern staging setup as @image01.

· Prompt: *"Replace the couch in @Mark01 with the modern sofa from @image01."*

· Generate. Staging without the stager.

For color changes: paint the wall as @Mark01, reference @Hex01 for "warm white", prompt to repaint. Buyers see a clean canvas instead of the previous owner's olive green.

4. For Small Business Marketing

You run a coffee shop. You want a banner ad for Instagram. You shot a photo of your latte, but the foam art looks weird and the cup has a chip on the rim.

With Image Canvas V1:

· Mark the chipped rim as @Mark01.

· Switch to Doodle mode and sketch a cleaner rim.

· Prompt: *"Clean up the rim in @Mark01 using my doodle."*

· Generate. Ship the banner. Same latte, polished look.

For batch edits: use the same prompt across multiple photos. Tag the same @Mark01 position on each photo. Get a consistent visual style across a campaign without a designer.

 

Viddo Image Canvas V1 vs The Alternatives

Image Canvas V1 isn't the only AI image editor on the market. Here's how it compares to the tools most欧美 creators are already paying for.

Tool	Click Edit?	Prompt Length Needed	Doodle Mode?	Notes
Adobe Photoshop Generative Fill	Partial	Long (30-60 words)	No	Desktop only, steep learning curve
Canva Magic Edit	No	Long (20-40 words)	No	Easy to start, limited precision
Fotor AI	No	Long (20-40 words)	No	Prompt-driven, decent templates
Pixelcut	No	Long (20-40 words)	No	Mobile-first, good for product photos
Viddo Image Canvas V1	Yes	Short (5-15 words + @ tags)	Yes	Browser-based, no install

 

The differentiator isn't templates or polish. It's the interface. Image Canvas V1 is the only major AI image editor that lets you *compose* your edit visually — by marking regions, dropping reference images, and doodling — instead of describing all of it in text.

If you've ever felt that AI photo editors were built for prompt engineers, not for creators — Image Canvas V1 is the alternative.

 

Frequently Asked Questions

What does Viddo AI Image Canvas V1 do?

Image Canvas V1 is an in-browser AI photo editor that lets you mark regions of an image, doodle new shapes on top, and reference uploaded photos or color swatches with simple tags (@Mark, @image, @Hex). The editor turns these visual tags plus a short text prompt into precise edits.

Do I still need to write a prompt?

Yes — but a much shorter one. The Prompt Editor is right in the right sidebar, and the placeholder reads *"Type @ to insert tags, e.g. replace the sofa in @Mark01 with @image01…"*. That seven-word sentence is a complete prompt. Most Image Canvas V1 edits use 5-15 words, because the @ tags carry the visual context the model needs.

How do @ tags work?

When you paint a region in Mark mode, it becomes @Mark01 automatically. Uploaded reference images become @image01, @image02, etc. Color cards become @Hex01, @Hex02, etc. In the Prompt Editor, typing @ brings up a list of available tags. Drop one in, and the model understands what you mean visually — no description required.

Can I use the edited images for commercial projects?

Yes. Edits you create with Image Canvas V1 are yours to use commercially, including product photos, marketing materials, social media content, and client work. Just don't use them to break the law or impersonate real people.

What image formats does Image Canvas V1 support?

JPG, PNG, and WebP. Maximum file size depends on your plan, but the free tier comfortably handles typical product photos and social media images.

How is Viddo different from Photoshop generative fill?

Photoshop's Generative Fill is excellent but it's a desktop app with a steep learning curve, and it asks you to write detailed prompts of 30-60 words. Image Canvas V1 runs in your browser and lets you skip most of the description by using visual tags. For most creators doing everyday edits, Image Canvas V1 is faster.

What's the difference between Mark mode and Doodle mode?

Mark mode is for selecting a region you want to change — paint over it, get @Mark01, reference it in your prompt. Doodle mode is for sketching what you want to add — draw a new lamp, the model paints a real lamp. Use Mark to define *where*, use Doodle to define *what*.

Can I use Image Canvas V1 for batch product photos?

Yes. The Mark and tag system is designed for repeatable edits. Once you build a prompt for one photo (e.g., "Replace the background in @Mark01 with @image01"), you can apply the same edit to dozens of photos by marking the same region on each one.

Will my uploaded images be used to train AI?

No. Uploaded images are processed for the edit and deleted within 24 hours. They are not added to any training dataset. Your photos stay yours.

What aspect ratios and resolutions are supported?

Aspect ratios: 1:1, 3:4, 4:3, 16:9, and Custom. Resolutions: 1K (default), 2K, and 4K. Choose the ratio that matches your destination (Instagram square, YouTube thumbnail, vertical story, etc.).

Does Image Canvas V1 work on mobile?

Yes. The interface is responsive and works in mobile browsers. For detailed Mark edits, a tablet or stylus gives the best experience, but the tool is fully functional on phones.

Can I undo or revise an edit?

Yes. Every edit keeps your full history. Use Undo and Redo for small tweaks, or jump back to any previous version in the History tab. Each version keeps the original image, your marks, and your prompt.

 

Try Viddo AI Image Canvas V1 Today

If you've made it this far, you're probably the kind of person Image Canvas V1 was built for: someone who wants AI photo editing to be *easier*, not harder.

This is just V1. We're already working on more canvas features — multi-layer edits, background swaps as a one-click action, and collaboration so your team can mark up the same image together. The era of writing prompts like code is ending. Welcome to click to edit, doodle to create.