Questions, answered
Everything about models, shoots, credits, and campaigns.
Product
What exactly is Flashlook?
What exactly is Flashlook?
Flashlook is an AI fashion studio. You build a persistent virtual model once, then shoot that same model with any item - clothing, bags, jewelry, fragrance and more, from a flat lay, hanger shot, or phone photo - and generate catalog, editorial, lifestyle, and outdoor stills plus short on-model video clips without booking a photographer or model.
Who is Flashlook for?
Who is Flashlook for?
Independent sellers, DTC brands, boutiques, Shopify merchants, and creative teams who need on-model photography and product video at catalog speed. If you ship new SKUs weekly and can't afford a shoot per product, Flashlook replaces the studio day with a consistent AI model you control - in the browser Studio, via MCP agents, or inside Shopify Admin.
Is there a Shopify app?
Is there a Shopify app?
Yes. Flashlook AI is live on the Shopify App Store. Install it, sync your store catalog into Studio, generate on-model stills and short video on reusable fashion models, and push finished looks back to product media without leaving Admin. Start on the free plan, then upgrade via Flashlook billing when you need more credits.
How does Flashlook work with my Shopify catalog?
How does Flashlook work with my Shopify catalog?
After you install the app, sync products into Studio (new and updated listings keep syncing). Cast or pick a locked-identity model, shoot on-model looks from your flat catalog images, then push selected photos or clips straight into Shopify product media. Details and install link are on the Shopify page.
Do I need photography or retouching skills?
Do I need photography or retouching skills?
None. If you can upload a product photo and pick options from a menu, you can produce studio-grade on-model stills and short video clips. No Photoshop, no set building, no casting.
Is there an API?
Is there an API?
Developer API access is on the roadmap for Volume and Enterprise customers - generate at scale from your PIM, storefront, or internal tools. Join the waitlist by contacting us with your use case.
Models
How does the model stay consistent across shoots?
How does the model stay consistent across shoots?
When you lock a model we freeze an identity fingerprint - face, skin, hair, body - and reuse it as the reference for every generation. The same persona appears in every still and video, pose, and scene. No re-rolling faces between products.
How deep is the model builder?
How deep is the model builder?
Very. Monk skin-tone scale, hair color and style taxonomy, face structure, body type, height, age, origin presets, freckles, beauty marks, and more - hundreds of dimensions combined into trillions of unique looks. The live preview shows the real render, not a cartoon avatar.
Studio
What kind of product photos can I upload?
What kind of product photos can I upload?
Anything you already have: flat lays, mannequin shots, hanger photos, even iPhone captures on a bed or rack - and it's not limited to clothing. Bags, jewelry, fragrance, and other accessories work too. You don't need professional lighting - Flashlook reads the product and shoots it on your model with realistic fit and presentation.
What can I control in a shoot?
What can I control in a shoot?
Pose, expression, indoor or outdoor scene, studio backdrop color, weather, framing, lighting mood, editorial concept, props, and aspect ratio - plus still or video output in Studio. Every plan also includes Smart Pick (AI scene suggestions) and Auto-generate (one-click full shoot).
Can I generate video, not just photos?
Can I generate video, not just photos?
Yes. In Studio you can animate a look into a short on-model clip (a few seconds) with the same locked identity as your stills - for PDP hover media, ads, Reels, and Stories. Pick start and end frames or let Flashlook generate them, choose a video agent, and generate. Credits are charged per clip based on the agent and length.
What is Link2Shoot?
What is Link2Shoot?
Paste a product URL from any supported store. We pull the product image - clothing, bags, fragrance, accessories, whatever's on the page - and generate an on-model shot on a fresh AI model, instantly. The credit cost is shown before you run it, and you can try a couple of shoots free before creating an account.
How long does a shoot take?
How long does a shoot take?
Most single-item studio stills complete in under a minute, and Link2Shoot runs are even faster. Short video clips take longer - typically a few minutes depending on length and agent - and you'll see live progress while they render. HD still generations also take a few minutes for sharper detail.
What are Smart Pick and Auto-generate?
What are Smart Pick and Auto-generate?
Smart Pick suggests pose, scene, and styling combinations that fit your item. Auto-generate runs a full shoot with minimal input - upload in, polished image out. Both are included on every plan, free accounts included.
Credits & plans
Does the model builder cost credits?
Does the model builder cost credits?
Yes. Exploring the builder is free; generating the actual preview image costs credits. Model previews always render on our best all-round model so every model you save looks consistent, and the exact credit price is shown on the Generate button before you commit.
Can I save models on the free plan?
Can I save models on the free plan?
Yes - Free includes 1 saved model so you can reuse your first face. Starter includes up to 2 saved models; Brand includes up to 5; Boutique includes unlimited.
How do credits work?
How do credits work?
Credits pay for generated stills and video clips. Still prices depend on which image model you pick - from 40 credits on Bolt to 300 on Lumen High - with the cost shown in the picker and on Generate. Video clips are priced by agent and duration. Every image and video agent is available on every plan, including free; credits are the only limit. Paid plans refresh your balance every month (credits don't roll over).
What do I get for free?
What do I get for free?
600 welcome credits when you sign up - enough to preview a model and run a handful of test shoots. Every image model is available on every plan; what an image costs in credits depends on which one you pick (the fastest is the cheapest, the most detailed costs the most). Smart Pick and Auto-generate are included too. Free accounts don't include saved models, and credits don't refresh monthly. Upgrade when you're ready to ship a catalog.
Can I use the images commercially?
Can I use the images commercially?
Yes on paid plans. Starter, Brand, and Boutique include full-resolution stills and video with commercial license for your store, ads, and social. Use your welcome credits to evaluate quality before subscribing.
Campaigns
What is Campaign Builder?
What is Campaign Builder?
Bulk studio for boutiques: upload an entire collection, pick one or more models, define shots per item - front, back, profile, lifestyle - and generate a full indoor/outdoor campaign in one batch. Available on demand; contact us to scope production for your collection.
Do you offer done-for-you campaigns?
Do you offer done-for-you campaigns?
Yes. Our team can produce a full season - ghost mannequins, multi-model looks, indoor and outdoor - with two revision rounds on every deliverable. Contact us through the Enterprise form on the pricing page for a custom package quote.
Developers
Do I need a paid plan to use the API or MCP?
Do I need a paid plan to use the API or MCP?
MCP for ChatGPT, Claude and Cursor works on every plan, Free included. The REST API is included on every paid plan, from Starter at $19/mo. Keys are self-serve and spend the same credits as the studio. Rate limits grow with the plan (Starter 300 requests/min per key, up to 2,400 on Flex). The free plan covers the studio and MCP: 600 signup credits to try the workflow.
Does generating through the API cost more?
Does generating through the API cost more?
No. A shoot costs the same credits whether it comes from the studio, a script, or an agent. Reading options, listing your library and quoting a price are all free; only generation spends. Call the price-quote endpoint before a batch and you will know the exact cost first.
What can someone do with my API key?
What can someone do with my API key?
A key acts as you: your products, your models, your credits. Treat it like a password. You can create a separate key per machine or per CI job and revoke any of them independently, and a revoked key stops working on its very next request.
Why does a shoot return a task id instead of an image?
Why does a shoot return a task id instead of an image?
Generation takes 60 to 150 seconds depending on the render tier, which is far longer than a sensible HTTP request. So every generating endpoint answers immediately with one task id per image, and you poll the task endpoint every second or two until it reports succeeded or failed. Each image bills and refunds independently, so one failure in a batch of ten only costs you that one.
Can my agent shoot a whole collection at once?
Can my agent shoot a whole collection at once?
Yes. The Collection Autopilot tools take every photo of a drop and group them into products and colorways. They style each one into a full look and run the shoot two products at a time. The agent checks progress and hands you the image URLs. Starting a run requires the quoted credit total, so an agent can't spend before you've seen the price. The same flow is available over the REST API under /collections.
Which agents can connect?
Which agents can connect?
Anything that speaks MCP over HTTP and lets you set a request header: Claude Code, Cursor, Codex, VS Code, OpenClaw, Hermes. Add the connector URL with an Authorization header carrying your API key and the tools appear in that session.
What can an agent actually do?
What can an agent actually do?
Everything the API can. It can add products from local files, cast and render a fashion model, run a shoot and hand back image URLs, animate a still into a clip, browse what you generated last week, and check your credit balance before it spends anything.
How does an agent know the valid poses and scenes?
How does an agent know the valid poses and scenes?
It asks. The options endpoint returns every render tier, limit and taxonomy value with no key required, so an agent discovers what is valid instead of guessing. Unrecognised values are ignored by the pipeline rather than rejected, so responses also echo any body fields that were dropped - a typo shows up instead of quietly producing the wrong photo.
Are there rate limits?
Are there rate limits?
Yes, per key, and generous enough that a normal integration never notices. Credits are the real quota; the request limits exist so a runaway loop cannot saturate the render queue for everyone else. Responses carry standard RateLimit headers, and a 429 tells you when to retry.
Is there an OpenAPI spec or an SDK?
Is there an OpenAPI spec or an SDK?
There is a full OpenAPI 3.1 document you can point a generator or an agent at, linked from the API page. There is no official SDK yet - the API is plain JSON over HTTPS with a bearer token, so a few lines of fetch or requests is genuinely all it takes.
Still stuck? See pricing · Contact us

