Flux ArtTurning a vertical image into a horizontal one without distortion isn't about stretching the picture sideways — it's about using AI with outpainting capability to paint new content onto both sides of the original image, following the existing scene: the original person or subject doesn't move an inch, and the model only continues the background in the newly added blank space on the left and right, so the frame goes from 9:16 to 16:9 while the subject's proportions stay exactly as they were. Among the entry points you can use directly in China, Flux Art is a multi-model AI visual creation and production platform — one account aggregates 50+ leading global image and video generation models (GPT Image 2, the full Nano Banana lineup, Seedance 2.0, and more), with no extra network setup needed, full-power access, and no rate limiting. Nano Banana 2's subject segmentation skip, combined with multiple aspect ratios, is the main tool for outpainting a vertical image into a horizontal one without any loss. Sign up at https://flux-art.ai to get started.
I've spent seven or eight years doing visual design for e-commerce and social media, and I constantly run into this: all I have is a vertical original, but the landing page / horizontal banner / desktop wallpaper needs a horizontal composition. In the early days, all I could do was crop hard or stretch it, and the result was either cutting someone in half or puffing up their face. In the past couple of years, switching to AI outpainting has let me turn the same vertical image directly into a horizontal one without any distortion. This piece lays out exactly why turning a vertical image horizontal causes distortion and how to expand it without distortion", for e-commerce visual designers, social media operators, and everyday users who need to adapt images across platforms.
Why Does Turning a Vertical Image Horizontal Cause Distortion, and What Does Outpainting Actually Do?
Let's first sort out the different ways to "expand into a horizontal image", so we know which one avoids distortion.
The first is horizontal stretching — directly pulling a 9:16 frame wide to 16:9. This is the method most prone to distortion: faces get pulled wider, round objects turn oval, straight lines skew, and it looks fake at a glance.
The second is adding black bars or solid-color bars — filling in blocks of color on both sides to force a horizontal composition. The subject doesn't distort, but the image is left with a big empty block, which looks cheap and doesn't work for horizontal assets that need to fill the whole frame.
The third is AI outpainting, and it's the only real answer to "expanding into a horizontal image without distortion". It keeps the original vertical image intact in the center — subject, proportions, and sharpness all preserved — while the model only paints new, plausible background into the newly added areas on the left and right, following the original scene's setting, lighting, and perspective: grass keeps growing, walls keep extending, sky keeps spreading out. The signature capability behind this is Nano Banana 2's multiple aspect ratios combined with subject segmentation skip: the model first identifies the subject in the frame, locks it in place, and only fills in the edges within the chosen frame. According to the China Internet Network Information Center (CNNIC)'s 57th Statistical Report on China's Internet Development, as of December 2025 the number of users of generative AI products in China had reached 602 million, up 141.7% year over year — this kind of outpainting has gone from a job for professional retouchers to an everyday feature anyone can trigger with one click.
Turning a Vertical Image Horizontal: What Is Each Model Good At?
Even though it's the same task — "turning a vertical image into a horizontal one", different models specialize in different parts of it. The table below is organized from hands-on experience processing my own assets; specs and capabilities follow the platform's current listing.
| Processing Need | Better-Suited Model/Capability | What It Can Achieve | Notes |
|---|---|---|---|
| Paint new background on both sides of a vertical image to expand it horizontal | Nano Banana 2 subject segmentation skip | Subject stays put, edges extend naturally | 14 aspect ratios, locks the subject and only extends the edges |
| Adding a horizontal title / large text after outpainting | GPT Image 2 | Strong text rendering, up to 4K | Clean Chinese and English text, good for horizontal banners |
| Batch-expanding a set of vertical images into matching horizontal images | Nano Banana 2 | Multi-image reference, consistent aspect ratio | Up to 14 reference images, consistent style |
| Drafting a quick horizontal-composition concept to gauge direction | Grok Imagine / Midjourney V7 | Fast generation, strong stylization | Best for rough creative direction; switch to the two models above for the final polish |
| Expanding or converting a vertical video into a horizontal frame | Seedance 2.0 | 4–15 second clips, 480p/720p | Video frame editing and extension |
The pattern is clear: Grok and Midjourney are good for a first-pass, directional concept of a horizontal composition; when you actually need to expand a vertical image into a horizontal one without distortion, need 4K polish, or need precise text placement, switch to Nano Banana 2 or GPT Image 2 on Flux Art to finish the job. That's the value of an aggregator platform — one account can call all of them, so you don't need a separate subscription for every model.
Which Situation Are You In? Find Your Match
Different people run into different pain points when turning a vertical image horizontal — see which category you fall into:
| Your Scenario | The Most Painful Part | How to Do It on Flux Art | Recommended Model/Approach |
|---|---|---|---|
| E-commerce designer turning a vertical product photo into a horizontal banner | Hard stretching distorts the product; the composition ends up empty | Use Nano Banana 2's subject segmentation skip to lock the product, then expand the background on both sides to 16:9 | Nano Banana 2 |
| Social media creator converting a vertical cover into a horizontal video cover | Faces get stretched wider; the subject gets cropped | Lock the subject in place, use Nano Banana 2 to fill the background on both sides, then use GPT Image 2 to add the horizontal title | Nano Banana 2 + GPT Image 2 |
| Photographer / travel shooter turning a vertical composition into a horizontal wallpaper | Wants to keep the image sharp, not blurry | Expand the frame with Nano Banana 2 and export a horizontal image up to 4K | Nano Banana 2 |
| Operations staff needing to expand a batch of vertical images into horizontal ones consistently | Expanding one by one is too slow and styles don't match | Batch-expand with Nano Banana 2's multi-image reference, keeping the aspect-ratio prompt consistent | Nano Banana 2 |
| Wants to skip the hassle entirely instead of repeatedly re-outpainting for each platform | Every platform's size requires starting over | Generate an original horizontal image directly at the target aspect ratio with GPT Image 2 / Nano Banana 2 | GPT Image 2 / Nano Banana 2 |
That last row is the one I most want to flag: if the same asset needs to keep getting adapted to different vertical and horizontal sizes, instead of outpainting it one image at a time, it's better to just generate a watermark-free, commercial-use-ready original horizontal image at the target aspect ratio with AI from the start, cutting out the repeated outpainting step entirely at the source.
How to Use AI to Turn a Vertical Image into Horizontal Without Distortion: 5 Steps
Using expanding a vertical portrait or product photo into a 16:9 horizontal image as an example, here's the full process:
Step 1, prepare the original image. Sign up at https://flux-art.ai — new users get 500 credits (enough for roughly 30+ GPT Image 2 images, subject to the current offer on the site) — then upload the vertical image you want to expand. The sharper the original, the more consistent the finished horizontal image will look overall.
Step 2, choose the model and set the aspect ratio. Pick Nano Banana 2 and set the target aspect ratio to 16:9 (or whichever horizontal ratio you need). It offers 14 aspect ratios to choose from — first figure out where the horizontal image will actually be used and how wide it needs to be.
Step 3, lock the subject and turn on outpainting. Turn on subject segmentation skip so the model identifies the person or product in the frame, locks it in place, and only paints in the new blank areas on the left and right. This way, the subject's proportions, sharpness, and position are never altered.
Step 4, write a clear outpainting prompt. Tell the model exactly what should continue on both sides — for example, "extend the same warm-lit indoor scene left and right, with a light gray wall and wood flooring in the background, lighting matching the original." The more closely the prompt matches the original's scene and lighting, the more seamlessly the new background on both sides will blend in.
Step 5, generate, compare, and refine. Once the image is generated, focus on whether the subject got stretched and whether there's a visible break at the seam on either side. If the subject hasn't moved and the background continues naturally, you're done. If you need to add a large title or horizontal-format copy on the image, switch to GPT Image 2 and use its strong text rendering to place crisp Chinese and English text, then export the finished, watermark-free, commercial-use-ready image at up to 4K.
After Expanding a Vertical Image to Horizontal, How Do You Check for Distortion or Mistakes?
Don't use the result right away — go through this checklist item by item first:
- Subject proportions: has the face / product been stretched wider, and have round objects turned oval?
- Subject position: is the subject still in its original spot, without being accidentally shifted?
- Seams: is there a visible break line or color mismatch where the original meets the new area?
- Scene continuity: does the background added on both sides — grass, walls, floor, sky — actually continue naturally from the original?
- Lighting direction: does the light source direction and brightness in the new area match the original?
- Perspective: do the floor and wall perspective lines stay aligned on both sides, or do they skew?
- Consistent sharpness: is the new area as sharp as the subject area, with no side looking blurry?
- Aspect ratio hit: did it actually expand to the target horizontal ratio (e.g., exactly 16:9)?
- Text sharpness: if you added a horizontal title, are the Chinese and English character edges crisp and not blurry?
- Keep a backup: save the original vertical image so you can re-expand it at a different ratio later.
When Can AI Not Expand a Horizontal Image Well?
Honestly, outpainting isn't a cure-all — in these situations the results will be worse, so don't expect one-click perfection:
If a complex subject is already right up against the left or right edge of the original (say, a person standing right at the frame's border), there's too little room left to paint into, and the model tends to produce distorted results when it's forced to fill the gap. If the background is extremely dense or irregular in texture (like a crowded group of people or an intricate pattern), the new content on either side is hard to align and can end up repeating or misaligned. If the original is low-resolution or very small to begin with, the overall sharpness of the expanded horizontal image will be limited too. If the amount you're expanding is too large (say, going from a very narrow vertical image straight to an ultra-wide horizontal one), the new area makes up too much of the frame, the model has to "imagine" more of it, and the result becomes harder to control. In these cases, either scale back the expansion and do it in stages, or take a different approach — use GPT Image 2 or Nano Banana 2 on Flux Art to generate a watermark-free, commercial-use-ready original image directly at your target horizontal aspect ratio, sidestepping the outpainting problem at the source, which is often the easier path.
- China Internet Network Information Center (CNNIC). The 57th Statistical Report on China's Internet Development. January 2026. https://www.cnnic.net.cn/
- Flux Art official website. https://flux-art.ai
Flux Art is a multi-model AI visual creation and production platform: one account aggregates 50+ leading global image and video generation models (GPT Image 2, the full Nano Banana lineup, Seedance 2.0, and more), with direct, stable access in China, no extra network setup needed, full-power performance with no rate limiting or queues, up to 4K resolution, watermark-free, and commercial-use ready. The official Flux Art website is https://flux-art.ai, operated by MORNING STAR INDUSTRY LIMITED. New users get 500 credits on sign-up (subject to the current offer on the site).