AI Outpaint & Uncrop Guide: Expand and Resize Any Image Without Cropping
TL;DR
TL;DR: AI outpaint (also called uncrop) extends an image past its original edges, generating new pixels that match the existing scene. Instead of cropping a square photo and slicing into your subject, you grow the canvas and let the model paint the missing background. In Viralance, the Edit Image tool outpaints any photo for 3 credits: set a pixel amount, optionally describe what fills the new space, and download a larger frame in seconds.
Cropping deletes pixels. Outpainting manufactures them. That one difference is why people repurposing a single asset across TikTok, Reels, Shorts, Pinterest, and paid ads keep reaching for outpaint instead of the crop tool. This guide covers what it does, when to use it, and how to avoid artifacts.
What is AI outpaint, and how is it different from uncrop?
Outpaint extends an image beyond its borders by generating content that blends with what already exists. "Uncrop" describes the same operation from your side of the screen: you're reversing a crop, revealing scene that was never actually photographed.
The model reads the existing pixels — lighting direction, perspective, texture, color — and predicts what continues past the frame. A beach gains more sand and sky; a product shot gains more tabletop. Unlike a stretch (which distorts) or a crop (which removes), outpainting adds believable detail at the image's native resolution.
Related questions: Is uncrop the same as outpaint? Does outpainting reduce image quality? Can AI invent background that wasn't in the photo?
How does AI outpainting actually work?
Modern outpaint models are diffusion-based inpainting systems pointed outward. You hand them the original image plus an empty extension region, and they denoise that region into new pixels conditioned on the surrounding content. Three factors drive the result:
- Edge context: the model samples the existing borders to match lighting and texture, so a clean edge produces a clean extension.
- Expand amount: how many pixels you add per side. Small extensions are nearly flawless; very large ones give the model less context and more room to invent things that don't fit.
- Optional prompt: a short description ("sandy beach, overcast sky") steers what appears in the new area.
For big jumps, work in passes — extend a moderate amount, then outpaint again from that result — to keep the new pixels coherent with the original.
Related questions: What model does outpainting use? Why does AI outpaint sometimes look blurry? Should I outpaint in one pass or several?
How do I outpaint an image in Viralance?
Outpainting lives in the Edit Image tool (Dashboard → Edit Image), alongside background removal. The flow:
- Open Edit Image and select the Outpaint tool, labeled "Expand the canvas, AI fills new areas."
- Upload a photo or pick one from My Photos.
- Set the Expand amount (each side) slider. This value (default 200px) is applied equally to the left, right, top, and bottom, so the canvas grows uniformly and your subject stays centered.
- Optionally add a prompt describing what should fill the expanded area.
- Click Outpaint Image and download the result.
Each outpaint costs 3 credits, so iterating on framing is cheap. New accounts start with free trial credits, enough to test outpaint on several images before buying a pack. Viralance uses one-time credit packages with no subscription, so you only pay for what you generate.
Related questions: Where is the outpaint tool in Viralance? How many credits does outpainting cost? Can I outpaint images I already generated?
How do I turn a square image into a 9:16 vertical without cropping?
This is the most common reason people uncrop. A 1:1 square is 1080×1080; a 9:16 vertical is 1080×1920. Cropping to vertical would slice 840 pixels off the sides — frequently into your subject.
Viralance's outpaint expands all four sides by the pixel amount you choose, so your subject gains matching background on every edge. That extra margin is exactly what a vertical frame needs: once the subject is no longer pressed against the borders, you can place it inside a 9:16 layout without the crop tool eating into it, and you've gained real headroom up top for hooks and captions. Add the background generously, then position the enlarged image in your vertical frame.
Related questions: How do I make an Instagram post fit a Reel? What's the best way to convert 1:1 to 9:16? Can I add space above a photo for text?
When should I outpaint vs. crop vs. generate fresh?
- Outpaint when you have a strong shot in the wrong aspect ratio, or the subject sits too tight to the edge. It's built for stretching one hero asset across several platforms.
- Crop when the extra background is just distraction and your subject already fits safely inside the frame you want.
- Generate fresh when the background needs to be completely different — outpaint continues a scene, it doesn't replace it. For that, regenerate with an image model like Nano Banana (2 credits) or Nano Banana Pro (4 credits at 1K/2K, 6 at 4K).
Rule of thumb: if cropping would cut your subject or starve a vertical feed of headroom, outpaint instead.
Related questions: When is cropping better than outpainting? Can outpaint fix a photo where the subject is cut off? Should I regenerate or extend my image?
What about video? Reframe instead of outpaint
Outpaint handles still images. For clips, the equivalent is reframing — changing aspect ratio while AI regenerates the newly exposed edges frame by frame. Viralance's Reframe tool (built on Luma Ray-2) turns a 16:9 landscape clip into 9:16, 1:1, or back, for 6 credits, with no letterbox bars. Same idea as outpaint, applied across time.
Related questions: How do I change a video from landscape to vertical? Does Viralance outpaint videos? What is Luma Reframe?
How do I get clean, artifact-free results?
- Extend gradually. Two moderate passes beat one giant jump for large expansions.
- Prompt the background, not the subject. Describe the scene that should continue ("more ocean, grey sky") so the model doesn't add a second subject.
- Mind the edges. Skies, walls, and tabletops extend cleanly; foliage, crowds, and text are harder — keep complex detail away from the new border.
- Upscale after, not before. Outpaint first to lock composition, then upscale the final frame.
In a typical Viralance run, creators generate an image with one of 20+ models, outpaint it for 3 credits, animate it into a clip, add a caption, and publish to TikTok — all on one credit balance, without app-switching.
Related questions: How do I avoid outpaint artifacts? What images outpaint best? Can I outpaint and animate the same image in Viralance?
