To keep a location consistent across AI shots, generate one establishing image of the set first and feed that plate into every shot as a reference. A location is not one variable. It is four: geometry, light, materials, and dressing. Each drifts separately and each has its own fix, so diagnose which one moved before you touch the prompt. The product-catalogue version of the same problem, one room reused across every SKU, is in furniture product photography.
You lock the presenter. Every clip returns the same face, and it looks like a win. Then you cut the sequence together and the room is wrong. The window has moved to the other wall, the oak counter has turned walnut, the light that came from the left in shot one comes from behind in shot three, and a plant nobody asked for is now in frame.
Nobody notices a face that holds. Everybody notices a room that does not. This guide covers what actually drifts in a location, why a prompt cannot pin it, how to build a plate that every shot inherits, which models in the DesignerBox catalogue hold a set and how far, and what the whole thing costs in credits. It is written for marketing teams and agencies producing multi-shot campaigns, not evaluating whether to.
Key Takeaways
- A location is four variables, not one. Geometry, light, materials, and dressing drift independently. Fixing the wrong one wastes a generation.
- Diagnose before you re-prompt. Name which axis moved, then apply that axis’s fix. “Write a better description” is not a fix for any of the four.
- Words cannot pin a room. “A cosy kitchen with morning light” describes ten thousand kitchens. The model picks a different one each run. Show it the room instead.
- Build the plate once, then inherit it. One establishing image of the set, generated for 5 credits, becomes the reference every later shot and clip is anchored to.
- Reference budgets differ by model. Kontext Multi takes up to 10 references, Seedance 2.0 up to 9, Kling 2.6 Pro up to 4, and Veo 3.1 up to 3. The budget decides how many axes you can anchor at once.
- Batch the same location together. Shots generated in one pass against one plate hold better than the same shots generated across three sessions a week apart.
- The benchmarks separate these too. FilmBench scores environment consistency, lighting, and background layout as distinct dimensions, and finds weaker models fail to hold a consistently lit environment across shots (arXiv, July 2026).
What is location consistency in AI generation?
Location consistency means the same place reads as the same place across every shot in a campaign. The room keeps its layout, the light keeps its direction and temperature, the surfaces keep their material, and the props stay where they were. It is measured across shots, not inside one, so a clip that looks perfect alone can still fail the moment it sits next to the shot before it.
That cross-shot framing is how the research community treats it as well. The VideoMemory benchmark splits multi-shot consistency into character-persistent, prop-persistent, and background-persistent scenarios across 54 cases, and notes that current models “often fail to preserve entity identity and appearance when scenes change or when entities reappear after long temporal gaps” (arXiv, January 2026). Background gets its own category because it fails on its own terms.
Why does a location drift when a character holds?
A face is one object with one set of proportions. Anchor it and the model has one thing to keep still. A location is an entire frame of information, and every part of it is negotiable on every generation.
FilmBench, published by Alibaba Group, Moku Lab, and the Beijing Film Academy, evaluates a scene across separate sub-metrics for environment, foreground and background match, and tone and colour, and reports that “the gap opens further down the ranking, where weaker models struggle to keep a coherent, consistently lit environment across shots” (arXiv, July 2026). Degradation concentrates on cross-shot spatial consistency. That is three named failure surfaces for one location, from one benchmark, and it maps closely onto what goes wrong in production.
In practice a location drifts on four axes. Learn to name the one that moved and the fix becomes obvious.
Geometry: the room reflows
The window jumps to another wall. A doorway appears. The counter runs the other direction, or the camera is suddenly two metres further back than the shot it cuts from. Geometry is the axis viewers catch fastest, because a room that reflows breaks the sense that the shots were filmed in one place.
The fix: anchor to a plate that shows the spatial relationships, and keep camera language identical across shots. Change the subject and the action in the prompt, never the room description.
Light: direction and temperature move
Shot one has soft light from the left. Shot four has hard light from behind, half a stop warmer. Each shot is defensible alone. Cut together, the sequence looks like it was shot across three days.
The fix: state direction, hardness, and colour temperature in the same words on every generation, and generate a location’s shots in one batch rather than across sessions.
Materials: surfaces quietly restyle
Oak becomes walnut. Matte paint goes eggshell. Marble veining redraws itself, which is the one clients spot in a hero image. Material drift is subtle per shot and obvious in a grid.
The fix: materials are the axis words fail hardest at. This one is reference-only. Feed the plate, do not describe the surface.
Dressing: props come and go
A bowl appears on the counter. The plant from shot two is gone by shot five. Signage rewrites itself. Dressing drift reads as carelessness rather than as a different room, which makes it the cheapest to prevent and the easiest to ignore. The providers treat it as its own thing too: OpenAI’s video guide describes a reference image as locking “character design, wardrobe, set dressing, or overall aesthetic,” listing dressing separately from the subject (developers.openai.com, July 2026).
The fix: keep the prop list short and frozen. Every prop you name is a thing the model can reinterpret, so name only the ones that carry the shot.
Build the plate before you build the campaign
The move that fixes all four axes at once is a plate: a single establishing image of the set, approved before any shot is generated, then fed as a reference into everything that follows.
Build it as a wide, well lit, neutral frame showing the spatial relationships and the key surfaces. No subject, or a placeholder one. This is the set, not the shot. In DesignerBox a styled scene runs 5 credits and outputs up to 4K, so iterating a plate fifteen times costs less than one second of flagship video. Spend your indecision here.
Then inherit it. Every product still, every on-model frame, and every clip takes the approved plate as a reference alongside whatever else the shot needs. The prompt stops carrying the room and carries only the subject and the action, which is the split that makes the room stop moving.
Two rules make the plate hold. Batch the shots for one location in a single pass, because a set generated in one sitting drifts less than the same set rebuilt on Thursday. And when a sequence runs long, export the strongest finished frame and use it as the plate for the next run, so the location carries forward from the version you approved. The same chaining logic that holds a face holds a room, and the identity-side version is in our guide to keeping a character consistent across clips. Both sit inside a wider taxonomy: the four ways AI video distorts separates them from the failures no reference can fix.
Which models hold a location, and how
Every current model in the catalogue supports reference anchoring, but the reference budget differs, and the budget is what decides how many axes you can pin at once. Three references means a character, a product, and a plate, with nothing left for a style board. Ten means you can anchor the room from several angles. These sit on one DesignerBox subscription, so the anchor is chosen per shot rather than dictated by whichever single tool you happen to pay for.
| Model | Provider | How you anchor a location | Best for |
|---|---|---|---|
| Kontext Multi | Black Forest Labs | Up to 10 reference images per generation, combining character, pose, and setting references in one result (designerbox.ai, July 2026) | Building the plate, and holding a set across a still campaign |
| Seedance 2.0 | ByteDance | Omni reference, up to 9 images plus video and audio inputs, with 15-second multi-shot output (seed.bytedance.com, February 2026) | Longer multi-shot takes that must hold one room |
| Kling 2.6 Pro | Kuaishou | Elements, 1 to 4 images where the selected subject can be a person, animal, object, or scene (app.klingai.com, July 2026) | Shots where the setting itself is the reference |
| Veo 3.1 | Up to 3 reference images, plus start and end frame control (ai.google.dev, July 2026) | Narrative shots needing matched audio | |
| Sora 2 Pro | OpenAI | One reference image anchors the first frame, locking “character design, wardrobe, set dressing, or overall aesthetic” (developers.openai.com, July 2026) | Starting a clip inside an already approved set |
| Runway Gen-4.5 | Runway | References carry a subject across separate shots. Runway does not publish a reference count for Gen-4.5 (checked July 2026) | The same set reappearing across a sequence |
Read that table as a budget, not a ranking. Kling is the one that names a scene as a selectable element outright, and Kontext Multi is the only place in the catalogue where ten slots make it comfortable to anchor a room from several angles and a subject at once. On a three-reference model the plate is competing for a slot with the character and the product, which is the real reason to build the location on the image side first rather than asking one video generation to hold all three.
What holding a location costs in credits
The economics of a plate are the whole argument for building one, because the cheap step and the expensive step sit at opposite ends of the workflow.
An image generation or edit is 5 credits. A styled scene is 5. So a plate you iterate fifteen times, until the geometry, light, materials, and dressing are all exactly what the campaign needs, costs 75 credits. That is inside the 112 credits on the free plan.
Video is where credits actually go. It is priced as credits per second times duration, and it is by far the most expensive operation in the product.
| Example clip | Credits |
|---|---|
| Seedance Pro Fast, 720p, 5 seconds | 150 |
| Kling Standard, 720p, 5 seconds | 225 |
| Sora 2, 720p, 8 seconds | 1,600 |
| Veo 3 with audio, 8 seconds | 6,400 |
Put the bottom row against a plan. Premium includes 2,500 credits a month, so a single 8-second Veo 3 clip with audio costs more than two months of it. Regenerating that clip because the kitchen counter changed material is a 6,400-credit correction to a problem a 5-credit plate would have prevented.
The order that follows: lock the plate on the image step, block the sequence on a fast model to confirm the room holds through the motion, then spend flagship credits once. AI video needs Premium or higher and the commercial licence starts at Pro, both on the pricing page, and the per-second maths is in our breakdown of what AI video costs.
Where locations still drift, and the fix
Even with a plate in place, specific failures recur. Each maps to one axis, and none is fixed by rewriting the prompt.
| What you see | Axis | The fix |
|---|---|---|
| The window or doorway moves between shots | Geometry | Feed the plate into every shot and freeze the camera language. Change only subject and action. |
| Shadows fall the other way after a cut | Light | Describe direction, hardness, and temperature identically, and generate the location’s shots in one batch. |
| Wood, stone, or paint changes character | Materials | Reference only. Remove the material adjectives and let the plate carry the surface. |
| Props appear or vanish across the sequence | Dressing | Cut the prop list to what the shot needs, then keep it word for word identical. |
| The room resets after a scene cut | Geometry and light together | Export the last approved frame and use it as the reference for the next clip. |
| A long take slowly warps the background | All four, compounding | Shorten the clip. Build the move across several linked clips instead of one long generation. |
| Two locations bleed into each other | Geometry | Anchor each location with its own plate and generate them in separate passes. |
The last two rows are the current frontier. Long single takes and rapid location changes are where cross-shot spatial consistency degrades in the benchmarks, and no reference budget removes that. Shorter clips anchored to one plate are the reliable shape today.
Where DesignerBox fits
Location consistency is a chain: build the plate, hold it across stills and clips, keep every asset in one place, and rerun the set for the next product or client. DesignerBox is built around that chain rather than one link of it.
Every top image and video model sits on one subscription, so the plate gets built on the model with the widest reference budget and the clip gets generated on the model that fits the shot. Kontext Multi combines up to ten references into one consistent result, the widest budget in the catalogue for anchoring a set alongside a subject. One workspace and one library keep the plate next to the shots it produced, rather than scattered across drives and tool exports, which is where a set gets lost between campaigns.
A saved workflow reruns the same location setup for the next drop, so the room you approved in March comes back in June without rebuilding it. That is the structural version of the argument in our piece on why AI brand assets drift and how to stop it, and the apparel version is in holding one look across a fashion drop. See the Video Studio for where images and video share one canvas, or Workflows for turning an approved set into a repeatable pipeline.
The honest limit: a plate makes a location strongly consistent, not perfectly so. Check every shot against the plate before it ships, the same way you would check a face.
The lifestyle scene builder workflow reuses one approved set across a batch instead of re-describing it per shot. For interiors specifically, the virtual staging agent and the real estate virtual staging prompts library cover the repeatable version.
FAQ
Why do AI models struggle with location consistency more than character consistency?
Because a character is one object and a location is a whole frame of negotiable information. Geometry, light, materials, and dressing each drift independently, so a single anchor has four things to hold instead of one. Benchmarks reflect this by scoring environment, background layout, and lighting as separate dimensions rather than as one score (arXiv, July 2026).
Can I keep the same location across separate AI shots?
Yes. Generate one establishing image of the set, approve it, then feed that plate as a reference into every later shot and clip. Keep the room out of the prompt and let it carry only the subject and the action. Batch the shots for one location where you can.
How many reference images do I need for a location?
One good plate holds a set for most campaigns. Where a sequence moves through a room, anchor two or three angles of the same set. Reference budgets cap this: Kontext Multi accepts up to 10 references, Seedance 2.0 up to 9, Kling 2.6 Pro up to 4, and Veo 3.1 up to 3.
Can I keep a character and a location consistent at the same time?
Yes, but it costs reference slots. On a three-reference model, a character plus a plate leaves one slot for everything else, so a shot that also needs a product gets tight. Build the location on the image side with a wider reference budget first, then carry the approved frame into video as a single anchor.
Why does my background change even when the prompt is identical?
Because an identical text prompt is still a description, and a description is reinvented on every run. “A sunlit loft with concrete floors” matches thousands of rooms, and the model picks a different one each time. Only a reference image removes the choice.
Does a location have to be a real place?
No. A generated set works the same way as a photographed one, because the plate is doing the anchoring either way. Build the room you want once, approve it, then treat that image as the location for every shot that follows.
How do I keep a location consistent across campaigns months apart?
Save the approved plate and the setup as a workflow, and keep both in the same library as the shots they produced. Rebuilding a set from a written description months later reintroduces every axis of drift at once, which is the failure a saved reference exists to prevent.
Sources
- FilmBench sub-metrics for environment, foreground and background match, and tone, plus cross-shot degradation in weaker models: (arXiv, July 2026)
- VideoMemory multi-shot benchmark and its background-persistent scenarios: (arXiv, January 2026)
- Veo 3.1 reference-image budget and start and end frame control: (ai.google.dev, July 2026)
- Sora 2 Pro reference image locking character design, wardrobe, set dressing and overall aesthetic: (developers.openai.com, July 2026)
- Seedance 2.0 omni reference inputs and multi-shot output length: (seed.bytedance.com, February 2026)
- Kling 2.6 Pro Elements reference count and selectable subject types: (app.klingai.com, July 2026)
- Kontext Multi reference count for combining character, pose and setting: (designerbox.ai, July 2026)
- Runway does not publish a reference count for Gen-4.5 (checked July 2026)
- DesignerBox credit costs, plan allocations and feature gating verified against live product configuration, July 2026
Model reference limits and consistency features verified from provider documentation (Google, OpenAI, ByteDance, Kuaishou, Black Forest Labs, Runway) and from the FilmBench and VideoMemory benchmarks as of July 2026. This category moves monthly; re-verify a model’s reference budget before you rely on it. Individual results vary.
