API Flow Overview

Step-by-Step Explanation

Initiate Upload
- Make a POST request to /v1/video
  
  NB: Here's the recommended video specifications to get fast results - 1080p HD, 23.976 fps, low bit-rate, H.264 Codec. The generated video will maintain the same quality and video specs as the uploaded video
- Receive upload_url and video_id
- The API automatically creates project, scene, and actor
Upload Video
- Make a PUT request to upload your video file to the provided upload_url
- Call the success_url when complete or call the failure_url if failed
- Receive a shot_id for tracking
- To display actor pickers or previews, make a GET request to /v1/shots/{shot_id}/actors to retrieve actors for the shot. Each actor includes thumbnail_url — a short-lived, pre-signed HTTPS URL for that actor’s scene thumbnail.
Monitor Processing (optional)
- Make a GET request to check status using /v1/shots/{shot_id}/status
- Track both video processing and AI training status

Initiate Audio Upload
- Make a POST request to /v1/audio
- Receive upload_url and audio_id
Upload Audio
- Make a PUT request to upload your audio file to the provided upload_url
- Call the success_url when complete or call the failure_url if failed

Start Generation
- Make a POST request to /v1/shots/{shot_id}/generate
  
  NB: The generated video maintains the same video specs (resolution, frame rate, codec, colorspace, etc.) as the uploaded video. If you upload a 4K video, the output is always 4K. Outputs above HD resolution (1080p) may be more expensive than HD. If you want HD outputs via the API, you’ll need to upload and train your AI model with HD footage.
- Include the audio_id and output_filename
- Optionally specify language and maintain_expression
Monitor Generation
- Make a GET request to check status using /v1/shots/{shot_id}/generate/{generate_id}
- Track the generation progress
Download Result
- Once complete, download the result using /v1/shots/{shot_id}/generate/{generate_id}/download
- Receive a download_url for the dubbed video