Render concurrency

Video rendering is a resource-intensive process. A video may take a few seconds up to a few minutes to render depending on the length, resolution, frame rate, and animations used.

We automatically scale our rendering infrastructure in response to demand in order to ensure that there is always sufficient processing capacity. By doing so, video rendering capacity is maintained at all times, even during peak periods when thousands of videos are processed concurrently.

Server instances

Every user is guaranteed a maximum number of server instances. For the Essential tier, the number of parallel instances is limited to 15, whereas for the Growth and Beyond tier, the number is limited to 50. This means that if you are a Growth user and are rendering 50 videos simultaneously, all subsequent video rendering requests are queued until one of the reserved server instances becomes available.

Multipart rendering

There is a maximum time limit of 15 minutes for each rendering task. Most users will find this to be sufficient. However, if you intend to render long segments of 1080p video, you may run into this limitation. In situations like these, it is possible to divide the rendering process into multiple concurrent processes, a process known as multipart rendering.

Split points

Multipart rendering is accomplished by specifying split points. As the name implies, these are the points in the video where you wish to cut the video. This allows multiple server instances to work on different parts of a video at the same time, thereby speeding up the rendering process. Once each part has been rendered, it will automatically be concatenated to produce the final video.

If you wish to speed up the rendering process even further, make sure that the split points are distributed efficiently throughout the video. For example, splitting between scenes is recommended as only one server instance will have to download the media assets (images and video clips) associated with that scene.

Considering that there is only a limited number of server instances available per user, it is essential that you determine the split points that are most appropriate for your setup. A ten-minute video could be split into ten parts, but that would use up a lot of server instances. Or you could break it up into four parts in order to minimize the number of server instances used per video. Please contact [email protected] for assistance in this regard.


There are a few requirements for multipart rendering. For the API to determine which assets need to be downloaded for each video part, it needs to know the duration of each asset ahead of time. The following requirements must be met:

  • The main composition's duration property must be specified.
  • The main composition's frame_rate property must be specified.
  • For every video and audio element, the duration property must be specified.
  • For every video and audio element, the loop property cannot be set to true.
  • Every multipart segment should be at least 60 seconds long.
  • The number of multipart segments are limited to 5.

Multipart rendering is an advanced feature that should only be used if you are running into the 15-minute rendering limitation during the processing of your video. Please contact [email protected] for instructions on how to enable it for your account.

Previous page
The modifications object
Next page