Skip to main content

Prompt Based Translation

✨ Introducing Prompt-Based Translation in Dubverse

Updated over 3 weeks ago

At Dubverse, we're always innovating to improve your video localization experience. Today, we’re excited to introduce a powerful new addition to our Translation Services: Prompt-Based Translation.



🆕 What’s New?

Until now, users could choose between two translation models:

  • Hawk – Known for its speed and versatility

  • Eagle – Ideal for projects requiring more refined translation

Now, we’ve added a third option: Prompt Translation. This feature gives users the ability to guide translations by providing context-specific instructions, resulting in more tailored translations.



🎯 How It Works

Let’s walk through the process with a sample use case—generating English subtitles for a Hindi video.

  1. Start a Subtitle Project: Begin by uploading your video and choosing the “Create Subtitles or Dubbing” option.

  2. Select Target Language: Choose English or Hindi as your target language—Prompt Translation currently supports only these languages.

  3. Choose Prompt Translation: Under Advanced Output Options, select “Prompt” from the Translation Service dropdown.

  4. Enter Your Prompt: Click on "Enter Prompt" and describe the tone, context, or target audience for the translation. For example, “Make it conversational and easy to understand for teenagers.”

  5. Click "Let's Sub/Dub" to begin processing.

💡 Pro Tip: Make sure your prompt is appropriate and clear. Inaccurate or unsuitable language may result in translation failure.



✅ Why Use Prompt-Based Translation?

  • Customizes output based on tone, audience, and context

  • Helps avoid generic translations

  • Ideal for marketing, educational, or storytelling content

With this update, you have more control over how your content is translated—ensuring it resonates with your audience.


Let's see a example below -> For one project of Hindi to English subtitles, we passed a prompt as follow - "Create subtitle for this video in simple language" and we can clearly see the difference in highlighted segments as per the prompt passed.


Did this answer your question?