Intelligent video restoration operator, based on a multimodal large model to enable intelligent removal of watermarks and subtitles from videos. Supports automatic detection and removal of unwanted content such as watermarks, subtitles, and scrolling subtitles in videos, and outputs the restored video file.

Operator ID: las_video_inpaint

Core capabilities #

• Supports multiple target removal scenarios: watermarks, subtitles, scrolling subtitles, and more.
• Intelligent detection based on a multimodal large model, with flexible and accurate target identification and removal processing: precise detection, general detection; pixel_replace (default), pixel_generate.
• Supports precise subtitle masking, preserving edge details.
• Supports stable processing of long videos.
• Automatically retains audio without additional operations.

Effect showcase #

Supported regions #

• Beijing: cn-beijing
• Shanghai: cn-shanghai
• Guangzhou: cn-guangzhou

Operator performance #

Input requirements #

Output requirements #

• Billing standards

  Sub-item	Description of billing standards
  Billing item	Billing usage is calculated based on the duration of the output video. If the output video frame rate exceeds 30 fps, the duration of the output video will be adjusted by the multiplier of the actual output video frame rate to the base frame rate (30 fps), rounded up.
  Billing type	Pay-as-you-go, unit: CNY/minute, billed hourly based on actual billing usage.
  Unit price	The unit price depends on the video detection and video restoration mode you select when using video restoration.

  warning

  • Total payment: In addition to the unit price, the service billing is related to the actual restored video duration and video frame rate.
  • Video restoration does not change the video frame rate, but the duration of the restored video may be shorter than the input video duration.

• Billing details
  Billing formula: Total cost = Unit price * Billing usage = Unit price * (Video duration * Duration conversion coefficient)

  Sub-scenario	Unit price	Billing duration conversion coefficient (per thousand tokens)
  Standard detection mode - pixel_replace restoration mode, that is: detection_precise_mode ：false inpainting_backend：pixel_replace	1.2 CNY/minute	The base billing unit price is based on the billing unit price for the base frame rate of 30 fps. If the output frame rate exceeds 30 fps, the duration of the output video will be adjusted by the multiplier of the actual output video frame rate to the base frame rate (30 fps), rounded up, and billing will be based on the adjusted output video duration.
  Standard detection mode - pixel_generate restoration mode, that is: detection_precise_mode ：false inpainting_backend：pixel_generate	1.6 CNY/minute
  Precise detection mode - pixel_replace restoration mode, that is: detection_precise_mode ：true inpainting_backend：pixel_replace	2.8 CNY/minute
  Precise detection mode - pixel_generate restoration mode, that is: detection_precise_mode ：true inpainting_backend：pixel_generate	3.2 CNY/minute

• Billing example
  When performing video restoration using the standard detection mode - pixel_replace restoration mode, if the actual output video frame rate is 50 fps and the video duration is 10 minutes:

  • The multiplier of the actual output video frame rate to the base frame rate is 1.6, which is rounded up to 2.0.
  • The actual cost is:
    Total cost = 1.2 CNY/minute * (10 minutes * 2) = 24 CNY.

Submit #

Interface description #

Call the video inpainting operator to submit a restoration task, supporting intelligent removal of watermarks, subtitles, logos, and other content from video source files on the public network, Volcano Engine intranet, or TOS.

Request parameters #

Response parameters #

Example #

Request example

# Set OUTPUT_TOS_PATH to a TOS path writable by this account
export OUTPUT_TOS_PATH="tos://path/to/output"

curl --location "https://operator.las.ap-southeast-1.volces.com/api/v1/submit" \
--header "Content-Type: application/json" \
--header "Authorization: Bearer $LAS_API_KEY" \
--data '{
    "operator_id": "las_video_inpaint",
    "operator_version": "v1",
    "data": {
        "video_url": "https://*****/watermark_video.mp4",
        "output_tos_path": "$OUTPUT_TOS_PATH",
        "targets": ["watermark", "subtitle"]
    }
}'

Response example

{
    "metadata": {
        "task_id": "task-20251125163544-abc123",
        "task_status": "PENDING",
        "business_code": "0",
        "error_msg": ""
    }
}

Error code #

Poll #

Interface description #

Query the execution status and result of the video restoration task.

Request parameters #

Response parameters #

Example #

Request example

curl --location "https://operator.las.ap-southeast-1.volces.com/api/v1/poll" \
--header "Content-Type: application/json" \
--header "Authorization: Bearer $LAS_API_KEY" \
--data '{
    "operator_id": "las_video_inpaint",
    "operator_version": "v1",
    "task_id": "task-20251125163544-abc123"
}'

Response example

{
    "metadata": {
        "task_id": "task-20251125163544-abc123",
        "task_status": "COMPLETED",
        "submit_time": "2026-07-21T16:04:06+08:00",
        "end_time": "2026-07-21T16:04:06+08:00",
        "business_code": "0",
        "error_msg": ""
    },
    "data": {
        "inpainted_video_path": "tos://bucket/output/video_inpaint/xxx/output.mp4",
        "subtitle_bbox": "120,800,1800,960",
        "resolution": "1920x1080"
    }
}

Error code #