On this page

AI Upscaling

Introduction

AI Upscaling is a feature that increases the resolution of a video using AI-powered upscaling models, adding detail and clarity beyond what standard resizing can achieve. Qencode analyzes the source video and reconstructs it at your target resolution using one of three quality tiers, then continues the result through the rest of your encoding pipeline.

Please note

AI Upscaling feature is currently only available for Encoder Version 2

Enable AI Upscaling

Please note

AI Upscaling feature works only with 'mp4', 'advanced_hls', 'advanced_dash' outputs.

AI Upscaling is enabled by adding ai_upscale: 1 to the output format in your transcoding job, along with a target width and/or height. Use the /v1/start_encode2 method to launch a transcoding job.

To select a quality tier, add the optional ai_upscale_mode parameter to the same format object.

Request Example

{
  "query": {
    "encoder_version": 2,
    "source": "YOUR_VIDEO_URL",
    "format": [
      {
        "output": "mp4",
        "ai_upscale": 1,
        "ai_upscale_mode": "standard",
        "width": 3840,
        "height": 2160,
        "destination": [
          {
            "url": "YOUR_STORAGE_URL",
            "key": "YOUR_KEY",
            "secret": "YOUR_SECRET"
          }
        ]
      }
    ]
  }
}

On the Transcode Media page, choose MP4, HLS, or MPEG-DASH as an output format.
Set your target Width and/or Height.
Turn on the AI Upscaling toggle.
Select a Mode to control quality and processing speed. Use Standard for general upscaling, Premium for low-light or noisy footage, or Max for the highest possible quality.

Output Settings

Parameter	Description
ai_upscale	Enables or disables AI upscaling for an output format. Accepted values: 1 - to enable, 0 - to disable. Default: 0.
ai_upscale_mode	Quality tier used for upscaling. Options: standard, premium, max. Default: standard.
width / height	Target output dimensions. At least one of the two must be set when ai_upscale is enabled.

Quality Modes

The Mode setting controls which upscaling model is used, balancing processing speed, footage type, and maximum output resolution.

Standard - General-purpose upscaling with the fastest processing time. Supports output resolutions up to 16K. Best for most footage where turnaround time matters.
Premium - Tuned for low-light and noisy footage, producing cleaner results in challenging conditions. Output resolution is limited to 8192px on the longest side.
Max - The highest available upscaling quality, at the cost of slower processing. Output resolution is limited to a total pixel count of 14,745,600 (equivalent to roughly 4K at a standard aspect ratio).

Resolution Limits

Mode	Constraint
standard	Up to 16K, no additional limit
premium	Longest side must not exceed 8192px
max	Total pixel count must not exceed 14,745,600

If your requested width and height exceed the limit for the selected mode, the job will fail validation before processing starts. Choose a lower resolution or switch to a mode with a higher limit.

Need More Help?

For additional information, tutorials, or support, visit the Qencode Documentation page Link or contact Qencode Support at support@qencode.com.