Video translation uses AI to convert video content into one or more languages through subtitle, speech, and lip-sync translation, delivering a natural viewing experience for global audiences.
Availability
|
Translation type |
Supported regions |
|
Subtitle translation |
China (Shanghai), China (Beijing), China (Shenzhen), China (Hangzhou), Singapore, US (Silicon Valley) |
|
Speech translation |
China (Shanghai), China (Beijing), China (Shenzhen), China (Hangzhou), Singapore, US (Silicon Valley) |
|
Lip-sync translation |
China (Shanghai), Singapore |
Advantages
Multiple languages and dialects
-
Supports 40+ languages and 10+ Chinese dialects.
-
A single task can generate outputs in 40+ languages simultaneously.
Multiple video formats
-
Supports MP4, WebM, MOV, and M3U8.
Customizable output
-
Supports MP3 and WAV audio output.
-
Offers personalized configurations for specific requirements.
Features
IMS supports subtitle, speech, and lip-sync translation. You can modify the machine translation results through post-editing.
|
Feature |
Description |
Highlights |
|
Subtitle translation |
|
Fast, accurate text translation for quick multi-language support. |
|
Speech translation |
Builds on subtitle translation with these additional steps:
|
Preserves the speakers' identities and emotions in the translated content. |
|
Lip-sync translation |
Builds on subtitle and speech translation by adjusting lip movements to match the translated audio. |
Produces highly realistic results where lip movements match spoken content, ideal for promotional or interactive content. |
|
Post-editing |
|
Adjust translation output to meet specific needs and achieve optimal results. |
Billing
Billing details are covered in Video translation billing.
Get started
IMS provides three methods to create and manage video translation tasks:
-
IMS console: A graphical interface for intuitive operations.
-
Cloud editing (Web): A refined editing workspace where you can import materials, translate content, and do post-editing.
-
OpenAPI: APIs for integrating video translation into third-party systems and automating batch processing.
Supported methods vary by translation type:
-
Subtitle translation: IMS console, cloud editing (Web SDK), OpenAPI
-
Speech translation: IMS console, cloud editing (Web SDK), OpenAPI
-
Lip-sync translation: IMS console, OpenAPI
Create a translation task
Method 1: Create in the console
-
Go to the Video Translation page in the IMS console.
-
In the upper-left corner, select a region as needed.
-
Click Create Translation Task.

-
Configure parameters:
-
Translation Method: Select Subtitle Translation, Speech Translation, or Lip-sync Translation.
-
Select Source File: Upload the video file to be translated. MP4, WebM, and MOV formats are supported.
-
Subtitle Source: Specify whether to erase the original subtitles and the subtitle source. The following methods are supported:
-
OCR: Extracts built-in subtitles from the video using OCR. Set the OCR Range to improve accuracy.
-
ASR: Converts speech to text when the video has no subtitles.
-
OCR/ASR: Tries OCR first, then falls back to ASR.
-
Specified SRT File: Loads an existing caption file to synchronize with the video.
-
All: Recognizes subtitles by ASR and refines them through OCR to correct typos.
-
-
Target Language: Select one or more target languages. The system generates corresponding output videos.
-
Storage Directory and File Name: Specify the storage location and name for the output file.
-
-
Click Submit Translation Task to create a task.
-
View the task status, configurations, and translation results in the task list. When the status is Processed, click View Details to view basic information, advanced settings, and outputs.

Method 2: Create using cloud editing
Preparations
If you are not familiar with cloud editing, start with its basic operations.
Procedure
-
Go to the Cloud Editing page in the IMS console.
-
In the upper-left corner, select a region as needed.
-
Click Create Editing Project, then click Edit in the Actions column to open the project.

-
On the Materials tab, click Import and select the materials to translate. After they are added to Materials, click the plus sign
or drag the file to the editing area.
-
Select the audio or video to translate in the editing track, then click AI translation. Configure the translation type, subtitle extraction, language, and subtitle erasure settings, and click Submit.

-
Wait a few minutes. The translated result appears in the track.

-
Click Generate As > Generate in the upper-right corner. Configure the video production parameters and click OK to export the translated video.

Method 3: Create through OpenAPI
-
Create a translation task
Call SubmitVideoTranslationJob to create a video translation task. Related parameters: Video translation parameters.
-
Query the result of a translation task
To get the status and result of a video translation task, call GetSmartHandleJob. The API returns processing progress, completion time, and output URLs.
-
Query translation task list
To view all ongoing or completed video translation tasks, call ListSmartJobs.
-
Delete a translation task
Call DeleteSmartJob to delete a completed task and release resources.
Post-editing (optional)
-
To correct translation output, enable post-editing before submitting the translation task.
-
For lip-sync translation, post-editing only supports modifying voice output. Lip movements cannot be modified.
Method 1: Use OpenAPI
Detailed API operations: Speech translation with post-editing.
Method 2: Use cloud editing (Web SDK)
Preparations
If you are not familiar with cloud editing, start with its basic operations.
Procedure
-
Go to the Video Translation page and select the task to modify.
-
Click Edit in the Actions column to open the project. Follow the video tutorial below to modify your translation results.
FAQ
How do I align subtitle timing with the audio waveform?
Example:
To split the translated subtitle "Great where are you" into two segments ("Great" and "where are you"), align the start and end times with the troughs of the waveform. This optimizes the post-editing effect for speech translation.

Word count limit during post-editing
Keep the adjusted content within 1.5 times the original word count. Exceeding this limit may cause overly fast speech.
Example
Initial translation: Let's talk about this later. We need to go home now.
Inappropriate adjustment: Let's discuss this matter in more detail at a later time. Right now, we should focus on heading back home as it's important to ensure we get there safely and in good time. We can revisit this conversation when we are both more relaxed and have ample opportunity to explore all the aspects thoroughly.
Appropriate adjustment: Let's pick this up another time. We should be going home now.
Reference
To submit translation tasks through OpenAPI, review the Video translation parameters first.