Translation Hub is a self-serve document translation service for organizations that translate a large volume of documents into many different languages. It's a fully-managed solution; you don't need to build or maintain your own web application or underlying infrastructure. After the initial setup, you can start using Translation Hub to get fast document translation at a cost-effective price.
Translation Hub leverages Google's Cloud Translation API and AutoML Translation, so you can reuse any of the resources that you created with those services, such as glossaries and custom translation models.
How does Translation Hub work?
For each translation request, Translation Hub completes a machine translation, preserving the layout and format of your documents. You can include a review step (advanced tier only) so that translators can review and edit the machine-translated content.
For an overview of the translation process, see Workflow overview.
The following benefits highlight how Translation Hub can improve the overall efficiency of your translation workloads:
- Self-serve translation
- Compared to traditional Translation Management Systems, which are process heavy and start with human translators, Translation Hub lets you use AI and augment results with human translators. As part of a post-editing step (advanced tier only), you can bring in your own human reviewers to edit documents that were translated by Translation Hub. By including both AI and human-editing capabilities, Translation Hub offers timely and cost-effective translations.
- Document translation
- You can directly translate documents such as PDF and DOCX files. Translation Hub preserves the layout and format of the original documents so that translated documents have the same look and feel as the originals. For example, paragraph breaks and headings are preserved.
- Simplified administration
- Administrators manage all Translation Hub resources from the Google Cloud console. For example, in the Google Cloud console, you can set up portals, manage which business users have access to each portal, and create translation glossaries.
- Also, when you add business users to a portal, they create or use their existing credentials to sign in to the portal. Business users can access their assigned portals; they don't have access to Google Cloud.
- Repeatable workloads and continuous improvement
- To improve the quality and efficiency of future translations, you can use a translation memory (advanced tier only). A translation memory saves your human-edited translations so that they can be reused to improve the accuracy of future translations. You can also import existing translations to build a translation memory.
- Your translation memory data is owned by you and stays with you. You can also export your data to Cloud Storage and use it, for example, as training data to build custom models in AutoML Translation.
- Straightforward page-based pricing
- Translation Hub charges you based on the number of translated pages. Translation Hub also works with AutoML Translation (advanced tier only) models. You can create or use existing custom models that are in the same project as Translation Hub. Custom model translations are included in the per page translation costs. Any costs that are associated with training custom models are charged separately by AutoML Translation.
Sample use case
Imagine that you are in an organization that offers various time-sensitive reports to customers all over the world. Because you have readers across many different languages, you require a translation service that can quickly turnaround translations and handle many different documents.
Translation Hub provides self-service document translation to over 100 languages with a simple, cost-effective page-based pricing. Because Translation Hub is self-service, you run translations when you're ready and the results are returned a few seconds later. You can repeat these workloads consistently with many different documents in any of the supported languages.
You can also bring in your own reviewers to refine translations (also known as post-editing). You can save post-edits to a translation memory to improve the accuracy of future translations. If you use AutoML Translation, you export your translation memory data to train custom models in AutoML Translation. Compared to the generic translation model, a custom model can improve translations for your domain-specific content.
Translation Hub offers a basic and advanced pricing tier, where the advanced tier offers additional features at a higher translation cost per page. The pricing tier is set at the portal level when they are created. For more information, see Compare basic and advanced tiers or Translation Hub pricing.
- View the Workflow overview