Yes, users can generate captions for their media when exporting projects in WeVideo using the AI caption generator. The generated captions are editable and can be translated into different languages.
Before you start:
- Create a multimedia creation project and add your media to it.
Generate captions in the WeVideo Multimedia creation space
Generate captions when exporting the project
After editing the project, click the Export button in the top-right corner of the video editor.
Adjust the export settings, select the Upload Captions or Auto-generate Captions options, and click Export to start exporting the project.
Generate captions for the exported project
Locate the exported project on the Exports page, then click it to launch it.
Click the captions option and click the Auto-generate Captions button.
Edit the generated captions
After the captions are generated, users can edit them from the exported project. Click the captions from the exports to examine the captions.
Click the captions to edit the content and adjust their timecodes.
Switch to the Style and Placement tab to adjust the caption font, font size, text color, and alignment.
Translate the captions
Click the translate option at the top of the captions section, select the Add Translation option, and select a language from the supported list.
Share and download the generated captions
When sharing exports with others via links or embed codes, viewers can view captions by clicking the closed captions button on the player to enable them.
When downloading exports, users can choose from the following options to download captions.
- Combine video and captions: Burn the captions into the video and download it as a single video file.
- Separate video: Download the video file without the captions.
- Separate the captions: Download the captions file in the VTT format without the video.
How accurate are the generated captions in WeVideo?
| Note: Because of inherent variations in audio quality, speaker traits, and environments, no speech-to-text service guarantees the accuracy of its output. While there are third-party leaderboards and rankings, these are based on test data, and real-world performance may vary. |
WeVideo uses the Whisper model for auto-captioning for Educational license (Instructor and Organization) users and Enterprise license users, and the AWS Transcribe service for Individual Creator or Teams license users. In the test environment, AWS Transcribe achieves a Word Error Rate of approximately 12–15% on clean, general-purpose English audio under optimal conditions, corresponding to roughly 85–88% word-level accuracy. The Whisper model achieves a similar result. Please note that results may vary depending on audio quality, language, and context.
To ensure accuracy, WeVideo provides an in-platform captions editor that lets users easily edit and correct generated captions when needed.
For more support and troubleshooting, visit our Help Center. New to WeVideo? Try for free or contact our team to learn more.