← Back to blog

How to Make a Real Estate Video with AI

A 4-step AI workflow for real estate listing videos: still photos to motion, exterior cuts, lifestyle b-roll, and voiceover. $4 in compute vs $400 traditional.

You can make a real estate listing video with AI in under an hour using still photos from your MLS package and three models: Seedance 2.0 for photo-to-video, Kling 3.0 for exterior and interior cuts, and Nano Banana Pro for polished stills you want to hold on screen. A 60-second listing video for a $650,000 3-bedroom house costs roughly $4 in model credits on 8frame. A traditional videographer runs $350 to $500 for the same deliverable, with a 3 to 5 day turnaround.

TL;DR

Why AI for real estate listings now

In 2024, motion artifacts on furniture and room geometry were too obvious: rooms stretched at the edges, windows blew out into white voids, mirrors doubled or smeared. Seedance 2.0 changed the calculus by animating from an actual reference photo rather than generating a new room that resembles it. The room stays the room. Furniture doesn't drift. It's not indistinguishable from drone footage, but it clears the quality floor buyers need at decision stage.

At scale, an agent running 20 listings a month is looking at $7,000 to $10,000 per year in videography. With AI that drops to under $100 in model credits. The agents getting the most from this aren't replacing luxury showcase films. They're video-enabling every listing, not just the ones with a commission big enough to justify a crew.

The 4-step workflow

Step 1: Animate listing photos with Seedance 2.0

Upload your MLS listing photos as reference images in 8frame. Seedance 2.0's photo-to-video builds motion from the reference image rather than inventing a scene from a text description. The model holds the actual room.

Prompt structure: [Room description] with [specific motion]. [Lighting]. Slow [direction] camera move, holds composition. No warping on walls or furniture. 16:9, 5 seconds.

Tested prompt, primary bedroom:

A spacious primary bedroom with light hardwood floors, white walls, and a king bed centered under a window. Slow dolly-in toward the bed from the doorway. Warm afternoon light through sheer curtains, soft shadows. Static furniture, no motion artifacts. No wall distortion on edges. 16:9 horizontal, 5 seconds.

Generated in 67 seconds. Dolly-in was smooth, bed framing held throughout. One variant had slight baseboard warping; the second was clean. Run 2 variants per room and pick the tighter one.

Tested prompt, kitchen:

A bright kitchen with white shaker cabinets, marble countertops, and stainless steel appliances. Camera pans left to right across the island, ending on the window above the sink. Natural daylight, no overhead light. Countertops stay flat throughout pan. 16:9 horizontal, 5 seconds.

Pan was smooth, countertop stayed flat, appliance reflections had no shimmer. Generated in 71 seconds.

Step 2: Exterior and supplemental cuts with Kling 3.0

For exterior shots and any angle your listing photos don't cover, Kling 3.0 handles wide-angle architectural scenes better than Seedance photo-to-video. Use it when generating from scratch.

Tested prompt, exterior approach:

Two-story craftsman house, white trim, covered front porch, set back from street behind a mature oak tree. Camera starts at end of driveway, slowly moves toward front door. Late afternoon golden hour light, long lawn shadows. House stays geometrically stable throughout. 16:9, 5 seconds.

Generated in 64 seconds. Tree shadow movement was a bonus that wasn't prompted. House geometry held cleanly on the push-in. The oak tree flickered at 3.5 seconds on one variant; the second was clean.

Tested prompt, living room hero angle:

A living room with vaulted ceilings, exposed wood beams, and a stone fireplace centered in the far wall. Large sectional sofa in the foreground. Camera elevated in the corner, wide angle, slowly descending and pushing in. Warm evening light from floor lamps, cool ambient light from the window. Proportions consistent throughout move. 16:9 horizontal, 5 seconds.

Vaulted ceiling held scale on both variants; fireplace stone texture stayed sharp throughout the descent. A shot type that normally requires a ladder or drone rig on a real shoot.

Step 3: Lifestyle b-roll

Lifestyle shots establish neighborhood context and daily-life imagination for buyers. These don't require the listing photos; you're generating supporting atmosphere, not room footage.

Tested prompt, neighborhood context:

A quiet suburban street in early morning. A couple walks a golden retriever along the sidewalk, dappled light through mature street trees. Slow pan following the couple from behind. Warm golden morning light, long shadows. 16:9, 5 seconds.

Motion was natural; tree canopy light flicker was the most cinematic frame in the batch. Generated in 62 seconds.

For listing photo cleanup (removing clutter, converting a dusk shot to daytime), Nano Banana Pro handles those still-image fixes at higher resolution than pulling a frame from video. Drop the enhanced still into your timeline as a hold frame.

Step 4: Voiceover or music and export

Voiceover reads better on Zillow listings where buyers are in decision mode. Ambient music plays better on Instagram where the video needs to stand on its own in a scroll.

A simple cut order for a 60-second version: exterior approach (0 to 4s), living room hero angle (4 to 9s), kitchen pan (9 to 15s), primary bedroom dolly (15 to 21s), additional rooms (21 to 38s), lifestyle b-roll (38 to 48s), exterior hold with agent contact (48 to 60s).

Export 16:9 at 1080p minimum for Zillow, Realtor.com, and email. Export 9:16 for Instagram Reels and TikTok. The 9:16 cut is a center-crop of the 16:9 timeline; most room footage holds the key subject in the center third, so the crop works without re-editing.

Routing by property type

Not every listing needs the same model mix.

Luxury ($1.5M+): Seedance 2.0 for all interiors, Kling 3.0 for architecture and exterior, Nano Banana Pro for hero stills. Run 2 to 3 generations per room. Budget $8 to $15. A real drone pass for aerials is worth adding at this tier; AI covers interior and lifestyle.

Mid-market and starter ($400K to $800K): The workflow above applies directly. Budget $3 to $6. Commissions at this tier don't support $400 videography on every listing. This is where AI earns its keep.

Commercial (office, retail, industrial): Kling 3.0 is the primary model. High-ceiling geometry and large open floorplates generate better with Kling than with Seedance photo-to-video. Lifestyle b-roll shifts to commuter and urban context instead of neighborhood walks. Budget $5 to $10.

Walkthrough example: $650K 3-bed house, 60-second listing video

Here's the exact run for a 3-bedroom, 2-bathroom craftsman in a mid-sized market at $650,000.

Inputs: 14 MLS listing photos covering exterior front, back, living room, kitchen, bedrooms, primary bath, backyard, and street view.

Generations run on 8frame:

Clip Model Cost
Exterior push-in Kling 3.0 $0.32
Living room descent Kling 3.0 $0.32
Kitchen pan Seedance 2.0 $0.58
Primary bedroom dolly Seedance 2.0 $0.55
Primary bath push Seedance 2.0 $0.55
Bedroom 2 hover Seedance 2.0 $0.55
Backyard pan Kling 3.0 $0.32
Neighborhood lifestyle Kling 3.0 $0.32
Enhanced exterior still Nano Banana Pro $0.08

Total: $3.59 in model credits. Edit time including export: 45 minutes. A local videography service runs $375 to $425 for the same deliverable with 3 to 4 days turnaround.

Pitfalls

Window blowout. Blown-out windows in the original listing photo animate into the generation. Fix: run the photo through Nano Banana Pro to recover the window exposure before uploading. Or use Kling with "window balanced, exterior visible through glass, no overexposure" in the prompt.

Room scale drift. Kling occasionally loses the furniture-to-wall proportion on tight dolly-ins in small rooms, making the space look like a dollhouse by the end of the clip. Add "furniture proportions consistent throughout, room scale stable" to the prompt and generate 2 to 3 variants.

Mirror artifacts. Any visible mirror generates unpredictably. The reflection either doubles the camera move or freezes as a still while the room moves. Keep mirrors out of the prompt framing when possible. If you can't avoid a large vanity mirror, keep the camera angle shallow so the mirror is edge-frame rather than center.

FAQ

Are AI listing videos allowed by the MLS?

Most MLS systems don't have rules against AI-generated video as of June 2026. The relevant requirement is accuracy: the video must represent the actual property. Generating rooms that don't exist or artificially expanding square footage would violate standard misrepresentation rules. The workflow here animates actual listing photos, so the rooms in the video are the actual rooms. Check your local MLS addendum; some boards are adding AI disclosure language in 2026.

Can I use my existing MLS listing photos?

Yes, and they're the preferred input. Upload the listing photo to the 8frame canvas and use photo-to-video mode rather than text-to-video. The model holds the actual room and animates from it. You get better results from an average listing photo than from a detailed text description of a room you don't have.

What aspect ratio works best for Zillow and Instagram?

16:9 at 1080p for Zillow, Realtor.com, and MLS embeds. Zillow's player is horizontal-native; a 9:16 video will letterbox and waste the viewport. 9:16 for Instagram Reels, TikTok, and Stories. Edit your 60-second master in 16:9 and export a center-cropped 9:16 version from the same timeline. Most room footage centers the key subject, so the crop works without re-cutting.


For the full model comparison across video types, the AI video for ecommerce complete 2026 guide covers the same model decision logic. The 8frame workflow library has a real estate listing template with timeline, LUT, and export settings ready to clone.

Related articles

use caseHow to Make a Shopify Product Video with AIuse caseHow to Make a UGC Ad with AI (Without Filming)use caseHow to Make a Coffee Ad with AI

Your frames start here

Watch the canvas power your creative flow in real time

Stay in the loop

Be the first to hear about our launch and get product updates