Photoshoot GPT: AI Fashion Photo & Video Generator

Turn your clothing images into AI-powered fashion photos and videos with Photoshoot GPT. Show your collection on models, create new versions of existing fashion photos, and add movement to product images for your online store, marketplace listings, and social media.

Built for clothing brands, online sellers, creative teams, and developers automating fashion content, this Actor offers four ways to create from your product and reference images. Run it with your Apify account; you do not need to supply a separate Photoshoot account or API key. Each run performs one selected operation and saves the original files and result records to that run’s private Apify storage.

Pricing: On-model photos $0.50 per pose; replacements $0.25 per image; videos from $1.25 per video. Charges apply to successful outputs. See all prices.

What you can create

On-model product photos

Create a new photo of a model wearing your clothing, starting with photos of the product. Choose a model, background, and compatible poses from the visual catalog to shape the presentation—from a clean catalog look to a styled scene for a collection.

Product photo + selected model and background → on-model photo.

Select 1–10 distinct poses in one run. Each selected pose requests a separate photo, so you can build several views of the same product without submitting each pose as a new run.

Model replacement

Create a new version of an existing fashion photo with a different model. Upload a reference photo that already features your clothing and select the model you want to use. Explore other model looks for the same outfit or develop alternatives for a campaign.

Product replacement

Use a reference fashion photo to imagine your own product in that setting. Supply your product photos and a reference image; the reference guides the scene and composition while Photoshoot generates a new presentation of your clothing.

Product videos

Turn product photos into a video using a selected catalog model and one of two motion options:

• Custom motion: describe the movement you want, such as a gentle turn toward the camera. Use motion_mode: "text" and provide motion_prompt.
• Motion template: choose an available movement from the motion catalog. Use motion_mode: "template" and provide motion_template_id.

Example fashion video. The corner image illustrates the look; video runs use product photos and a selected catalog model, not a reference-photo input.

These modes use different generation paths and have different prices. Choose the mode that fits the movement you want; the Actor does not accept a custom motion description and a motion template together.

Fashion content use cases

• Online store and marketplace listings: create on-model product photos and choose an image aspect ratio that suits your listing layout.
• Collection launches: explore a consistent selection of catalog models, backgrounds, and poses across your products. Each run handles one product or reference image operation.
• Campaign alternatives: try another model in an existing fashion photo or use a reference scene to explore a different product presentation.
• Social media and product pages: add motion to product photos without filming new footage.
• Automated content workflows: submit structured inputs through Apify and retrieve the saved files and result records for your own pipeline.

How to create your first fashion photo or video

Try the included example: the example input selects Use the included example photo ($0.25) and model replacement. Start it without uploading files to generate one real image using the Photoshoot demonstration photo and model. It uses the same production, private delivery and billing flow as other runs; it is not a free preview. This small example also supplies the daily Apify Store quality test.

For your own images, turn off Use the included example photo first. Then follow the steps below and explicitly confirm your own media rights. Example mode rejects custom images/selections rather than applying publisher consent to them.

1. Choose what to create. Turn off the included example, then select on-model photos, model replacement, product replacement, or video in the Actor input form.
2. Upload your images. Use photos of the same product for product-based operations. Replacement operations also need a reference fashion photo. Use JPEG, PNG, or WebP files, up to 10 MiB per input image. This limit applies to both product and reference photos.
3. Browse the Photoshoot visual catalog. Copy the active IDs for the selections your operation needs. For on-model photos, choose a product type, model, background, and compatible poses. For template video, also choose a motion template.
4. Allow access to your uploaded files. After uploading, select the key-value stores containing those files under Input file stores. For example, if your product and reference photos are in the same store, select that store once. The selection must match the source image URLs and grants this run read-only access.
5. Confirm your media rights. Review the terms and media rules and privacy policy. In Published media policy version, enter media-upload-policy-2026-08-29-v2. Confirm Media rights and consent only when you have the required rights and consent.
6. Review the price and run budget, then start. Set Maximum cost per run to cover all requested outputs. When the run finishes, open Output to find the saved files and job summary.

Operations and prices

Operation	Input value	Price per successful output
On-model photo	on_model_photo	$0.50 per pose
Replace a model	model_replacement	$0.25 per image
Replace a product	product_replacement	$0.25 per image
Video with custom text motion	video, motion_mode: text	$1.25 per video
Video with a motion template	video, motion_mode: template	$2.50 per video

On-model photos support 1–10 distinct catalog poses in one run. For example, three successful poses cost $1.50; a failed pose is not charged as a successful output. Other operations produce one output per run. Platform usage is included in these prices. There is no separate Actor-start or automatic Dataset-item fee. Check the price shown by Apify before starting.

Set Maximum cost per run high enough to cover every requested output. The Actor checks the requested total before submitting generation; it does not silently reduce the number of poses to fit the budget. A run budget is not a query of your Apify account credit balance.

If a run delivers two successful poses and one pose fails, the successful-photo charge is $1.00. Successful outputs can still be charged when another part of the job fails or the run is cancelled. A saved file and a confirmed charge are separate steps; an interrupted delivery or charge may require same-run recovery.

The examples below use placeholder catalog IDs and stores. Replace them with your active catalog selections and actual uploaded files before running.

Input examples

Every operation requires operation, policy_version, and media_rights_confirmed. Product-based operations accept 1–10 photos of the same product and an optional description. Image operations accept an optional aspect_ratio (default 9:16); video uses its motion-mode settings and does not accept that field.

What you want to create	Required image and selection fields
On-model photos	product_images, product_type_id, model_id, background_id, pose_ids
Replace a model	reference_image, model_id
Replace a product	product_images, product_type_id, reference_image
Custom-motion video	product_images, product_type_id, model_id, motion_mode: text, motion_prompt
Template video	product_images, product_type_id, model_id, motion_mode: template, motion_template_id

Create on-model photos in two poses

Upload your product photos, then replace the product type, model, background, and two pose IDs with compatible selections from the catalog. See pricing for multi-pose costs.

{
  "operation": "on_model_photo",
  "product_images": ["https://api.apify.com/v2/key-value-stores/AAAAAAAAAAAAAAAAA/records/product.png"],
  "input_stores": ["AAAAAAAAAAAAAAAAA"],
  "product_type_id": "00000000-0000-4000-8000-000000000001",
  "model_id": "00000000-0000-4000-8000-000000000002",
  "background_id": "00000000-0000-4000-8000-000000000003",
  "pose_ids": ["00000000-0000-4000-8000-000000000004", "00000000-0000-4000-8000-000000000005"],
  "aspect_ratio": "4:5",
  "policy_version": "media-upload-policy-2026-08-29-v2",
  "media_rights_confirmed": true
}

Replace the model in a fashion photo

Upload the existing fashion photo as reference_image and select the replacement model. This operation does not require separate product photos.

{
  "operation": "model_replacement",
  "reference_image": "https://api.apify.com/v2/key-value-stores/AAAAAAAAAAAAAAAAA/records/reference.png",
  "input_stores": ["AAAAAAAAAAAAAAAAA"],
  "model_id": "00000000-0000-4000-8000-000000000002",
  "aspect_ratio": "9:16",
  "policy_version": "media-upload-policy-2026-08-29-v2",
  "media_rights_confirmed": true
}

Present your product in a reference scene

Upload photos of your product and a reference fashion photo. Select the product type for your uploaded product; this operation does not require a separate catalog model or background.

{
  "operation": "product_replacement",
  "product_images": ["https://api.apify.com/v2/key-value-stores/AAAAAAAAAAAAAAAAA/records/product.png"],
  "reference_image": "https://api.apify.com/v2/key-value-stores/AAAAAAAAAAAAAAAAA/records/reference.png",
  "input_stores": ["AAAAAAAAAAAAAAAAA"],
  "product_type_id": "00000000-0000-4000-8000-000000000001",
  "aspect_ratio": "4:5",
  "policy_version": "media-upload-policy-2026-08-29-v2",
  "media_rights_confirmed": true
}

Create a video with custom motion

Upload product photos, choose a model, and describe the movement. The example requests a gentle turn; generated motion can vary.

{
  "operation": "video",
  "product_images": ["https://api.apify.com/v2/key-value-stores/AAAAAAAAAAAAAAAAA/records/product.png"],
  "input_stores": ["AAAAAAAAAAAAAAAAA"],
  "product_type_id": "00000000-0000-4000-8000-000000000001",
  "model_id": "00000000-0000-4000-8000-000000000002",
  "motion_mode": "text",
  "motion_prompt": "The model gently turns toward the camera, showing the outfit from the front and side.",
  "policy_version": "media-upload-policy-2026-08-29-v2",
  "media_rights_confirmed": true
}

Create a video with a motion template

Choose an active motion template from the catalog. Supply its ID instead of a motion description.

{
  "operation": "video",
  "product_images": ["https://api.apify.com/v2/key-value-stores/AAAAAAAAAAAAAAAAA/records/product.png"],
  "input_stores": ["AAAAAAAAAAAAAAAAA"],
  "product_type_id": "00000000-0000-4000-8000-000000000001",
  "model_id": "00000000-0000-4000-8000-000000000002",
  "motion_mode": "template",
  "motion_template_id": "00000000-0000-4000-8000-000000000006",
  "policy_version": "media-upload-policy-2026-08-29-v2",
  "media_rights_confirmed": true
}

Source files and supported selections

Use the input form’s file-upload controls. For API submissions, source file URLs must have the query-free form https://api.apify.com/v2/key-value-stores/<store-id>/records/<record-key>, and input_stores must select the stores containing those files. Do not include tokens in the URLs. Arbitrary public image URLs and signed URLs are not accepted; uploaded bytes and access permissions are verified before import.

Send only the fields for your selected operation. On-model photos use catalog poses; free-text poses, image edits, and retry-generation commands are not supported by this Actor. Video duration, resolution, and aspect ratio are not configurable input fields.

View and download your results

Open the run’s Output tab to find:

Output link	What you will find
Generated media manifests	The default Dataset, with one result record per stored output.
Job summary and recovery instructions	The OUTPUT record, including job status, completed/failed counts, and recovery information.
Private output file records	A JSON list of keys in the run’s key-value store. This API link lists records; it is not a file download screen.

To download a photo or video:

1. In Apify Console, open the run’s default Dataset and choose its Output files view.
2. Find the row whose file role is original and copy its record_key.
3. Return to the run in Apify Console and open its default key-value store from the run’s storage section. Open the record matching that record_key to view or download the file. Use this Console storage screen rather than the Private output file records API link above.

For file details, job status, and recovery information, see Advanced results and recovery.

Example results

The following synthetic examples show one successful on-model photo. IDs, file size, hash, and pricing version are placeholders; they do not point to real media. The Dataset contains one manifest per saved output.

Dataset item:

{
  "schema_version": 1,
  "delivery_id": "00000000-0000-4000-8000-000000000020",
  "result_id": "00000000-0000-4000-8000-000000000021",
  "lineage_id": "00000000-0000-4000-8000-000000000022",
  "job_id": "00000000-0000-4000-8000-000000000023",
  "customer_id": "00000000-0000-4000-8000-000000000011",
  "run_id": "RRRRRRRRRRRRRRRRR",
  "result_version": 1,
  "ordinal": 1,
  "event": "on-model-image-pose",
  "pricing_version": "example-pricing-version",
  "dataset_id": "DDDDDDDDDDDDDDDDD",
  "dataset_item_key": "result_00000000-0000-4000-8000-000000000021",
  "media": [
    {
      "role": "original",
      "store_id": "SSSSSSSSSSSSSSSSS",
      "record_key": "result_00000000-0000-4000-8000-000000000021_v1_original",
      "mime_type": "image/jpeg",
      "size_bytes": 1000,
      "sha256": "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa"
    }
  ]
}

The record_key identifies the original file in the run's private key-value store; the sha256 is its content digest.

OUTPUT record:

{
  "id": "00000000-0000-4000-8000-000000000023",
  "status": "completed",
  "completed_count": 1,
  "failed_count": 0,
  "results": [
    {
      "schema_version": 1,
      "delivery_id": "00000000-0000-4000-8000-000000000020",
      "result_id": "00000000-0000-4000-8000-000000000021",
      "lineage_id": "00000000-0000-4000-8000-000000000022",
      "job_id": "00000000-0000-4000-8000-000000000023",
      "customer_id": "00000000-0000-4000-8000-000000000011",
      "run_id": "RRRRRRRRRRRRRRRRR",
      "result_version": 1,
      "ordinal": 1,
      "event": "on-model-image-pose",
      "pricing_version": "example-pricing-version",
      "dataset_id": "DDDDDDDDDDDDDDDDD",
      "dataset_item_key": "result_00000000-0000-4000-8000-000000000021",
      "media": [
        {
          "role": "original",
          "store_id": "SSSSSSSSSSSSSSSSS",
          "record_key": "result_00000000-0000-4000-8000-000000000021_v1_original",
          "mime_type": "image/jpeg",
          "size_bytes": 1000,
          "sha256": "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa"
        }
      ]
    }
  ]
}

Here id is the job ID and results contains the delivered manifests. Output contents can differ for partial completion, cancellation, errors, or recovery.

Frequently asked questions

Do I need a Photoshoot account or my own AI provider keys?

No. Use your Apify account and Apify billing; no separate Photoshoot or provider key is required.

Where do I find model, background, pose, and motion template IDs?

Copy active IDs from the visual catalog. Follow the selection steps in Getting started; example IDs are placeholders.

What is the difference between custom motion and a motion template?

One takes a written movement description; the other uses a catalog template. See Product videos for choosing a mode and pricing for its cost.

Can I generate several products or all four operations in one run?

Use separate runs for separate products or operations. A single on-model run can include several poses; see On-model product photos.

Can I use a photo from any public URL?

Use uploaded files in authorized Apify storage. See Source files and supported selections for accepted URLs and permissions.

Are the results added to my Photoshoot website account?

Actor results are stored in Apify, without synchronization to your existing Photoshoot website account. Follow View and download your results to retrieve them.

Will every generated image exactly match my product or reference?

AI-generated results can vary. Review clothing details, fit, composition, and motion before using them in a listing or campaign.

What happens if my run budget is too small?

The Actor rejects an insufficient run budget before submitting generation. See Operations and prices for budgeting, and follow any instructions in OUTPUT.

What if the run stops before I get all my files?

Keep the run and its storage. Follow Interrupted runs and recovery to recover the existing work safely.

Advanced results and recovery

Result details

The Generated outputs view shows output order, generation type, and result IDs. The Output files view shows the record key, file type, size, and SHA-256 digest for each file. A video may also have a preview or poster; download the original record for the original video. Images are at most 20 MiB and video originals at most 200 MiB. Access to these records requires permission to the run’s storage.

For example, two delivered on-model photos produce two Dataset records, with each record pointing to its own original file. A Dataset record identifies the saved files; check OUTPUT as well for the job outcome and any delivery or billing issues. Temporary Photoshoot media URLs are not stored in the Dataset.

The job summary reports completed or partially_completed, the backend job ID as id, completed/failed counts, and result manifests. Cancellation has a distinct cancelled outcome. Backend failed or blocked jobs fail the Apify run; handled Actor errors write an error summary where storage remains available.

Interrupted runs and recovery

Keep the same run and its storage. Use Resurrect to recover its saved job and outputs. Starting another run creates a separate request; it is not a safe way to retry an uncertain submission or charge. Do not change the input of an interrupted run. An input/checkpoint mismatch stops recovery instead of starting replacement production.

Cancellation stops new production. If successful outputs were already saved, the same resurrected run can finish their delivery and billing after backend settlement. This does not restart cancelled generation. Uncertain billing is reconciled against Apify’s event counters; the Actor never blindly sends another charge for a pending intent.

Error details for API users

Handled errors write an OUTPUT summary containing:

• code and error.stage: the failure category and processing stage.
• job_id, operation_id, backend_status, completed/failed counts when known.
• job_started: true for a known backend job, false before submission, or null when the submission outcome is uncertain.
• exported_output_count and dataset_count_verified_at_failure: delivered-file manifest count and whether it was freshly read. An unavailable count remains null.
• billing.status, delivery/intent IDs and the last checkpoint stage. count_verified_at_failure: false means no fresh charge-counter verification was performed during error reporting.
• error.action: recovery instructions. A definitive backend rejection may also include safe backend_code and http_status fields.

BILLING_RECONCILIATION_REQUIRED means keep the run storage and use Resurrect or contact support. BACKEND_RECOVERY_REQUIRED can mean a response was lost after a command committed; keep the same run. MARKETPLACE_CLOSED or ACTOR_CONFIGURATION_ERROR requires publisher attention. INSUFFICIENT_RUN_BUDGET requires a higher run spending limit or fewer explicitly selected poses before submission.

Completed file records remain in the Dataset/KVS when a later step fails. After an abrupt platform termination or storage outage, saving an updated summary may be impossible; keep the existing storage and inspect the run log. The last recorded delivery or charge state is not proof of a fresh platform verification.

Privacy and support

Upload only media you may lawfully process, with consent from depicted adults. Do not upload minors, violent or explicit content. Results remain private to authorized Apify users. For support, provide the Apify run ID, job ID and error code; never share API keys or runtime tokens.

Visit the Photoshoot website for product information and the visual catalog for available selections. For Actor issues, contact the publisher through this Actor’s Issues tab with the run ID and error details. Keep source images and generated media out of public support reports unless you intend to share them.