The VOD Express API turns recorded media into localized media. You describe once what to do (which language is spoken, which languages to translate into and dub, which files to produce), then drop video or audio files into a folder, or point the project at an RSS or Media RSS feed. Each file is transcribed, translated, dubbed with AI voices and packaged into the outputs you asked for.

Quickstart

Dub your first video into Spanish in about ten minutes, with curl.

How it works

Projects, where the media comes from and where the results go: the ideas behind the API.

Examples

Ready-to-send configurations for dubbing, subtitles, audio-only and batch jobs.

API reference

Every endpoint, with an interactive playground.

​
What you can build

One MP4 per episode with Spanish, German and the original audio as selectable tracks, cloned from the speakers’ own voices over the original background. Example.
SRT or WebVTT files in the source language and any number of translations, or subtitle tracks muxed into the video. Example.
Podcasts and audiobooks dubbed into MP3 or M4A. Example.
Sync a season folder to S3 with the AWS CLI; every file is processed as it arrives, and subfolders are kept in the outputs. Example.
The project reads your RSS or Media RSS feed, dubs each new item as it appears, and publishes the same feed with the Spanish files at a URL your players and partners can read. Feeds.

​
The flow at a glance

1

Authenticate

Exchange your SyncWords API key pair for a bearer token. Authentication.
2

Create a project

POST /projects with a config: the source language, the translations, the voices and the output files. Configure a project.
3

Upload media

Put files into the project’s input folder, with the AWS CLI or through the API. Upload media. A project that reads a feed takes its media from the feed instead. Inputs.
4

Collect the results

Follow GET /projects/{id}/files until each file is done, then download its outputs. Monitoring.

​
Hear it

The same clip before and after: a two-speaker English dialogue (synthetic voices over a test pattern), dubbed into Latin American Spanish by Deepdub, which clones each speaker’s voice (voice: "auto", locale: "es-XL"). The original voices are removed and the rest of the audio kept. Made with the Quickstart’s config plus the locale; the sample is yours to try it with.

Before: English

After: Spanish dub

​
Before you have keys

What is supported is public: every language with its engines, the voices to start with, and every field of a project’s config in the Config reference. To get keys, see Get access. The words these pages use are in the Glossary. Base URLs, conventions and the machine-readable description of the API are in the API reference; where tokens come from and what to check before going live is in Environments.