What are Profiles?
Profiles are collections of configurations for the MT engines and LLMs used to translate your content in Smartling. They allow you to customize translation output based on your specific requirements and preferences.
Once configured, Profiles can be used to:
- Generate translations in the translation step of a machine translation workflow
- Provide translation suggestions to linguists working in the CAT Tool
- Deliver translations via Smartling's MT API
- Support instant MT integrations to display translations directly where needed
Understanding Profile Types
Smartling offers two types of Profiles, each designed for different kinds of translation providers:
MT Profiles
MT Profiles are used to configure machine translation engines. They allow you to:
- Translate using your preferred MT engine
- Configure provider-specific settings (such as formality levels for DeepL)
- Define language-specific rules
- Enable features like profanity masking (for supported providers)
When to create an MT Profile:
- Create an MT Profile for any traditional MT provider (Google Translate, DeepL, Microsoft Translator, Amazon Translate, etc.), custom trained MT engine, or for Copy Source-to-Target.*
- *Google Translation LLM is an exception. If using Google Translation LLM, you will need to create an MT Profile, not an LLM Profile.
Learn more in Creating and Managing MT Profiles.
LLM Profiles (RAG)
LLM Profiles provide advanced customization options specifically for GPT (OpenAI), GPT (Azure), and Google Gemini (Vertex AI). They support capabilities beyond MT Profiles that are specific to translating with LLMs, including:
- Custom translation prompts - Write instructions that tell the LLM how to translate your content
- Prompt tooling with RAG - Augment prompts with contextual information from your translation memory, glossaries, and Automated Style Guides
- Translation parameters - Adjust output creativity, repetition tolerance, and other factors that influence translation output
- Hallucination detection - Automatically flag potential translation issues
- Token management - Configure token limits for your selected model
When to create an LLM Profile:
- Create an LLM Profile for GPT (OpenAI), GPT (Azure), or Google Gemini (Vertex AI) providers.
Learn more in Managing LLM Profiles and Prompts.
View the complete list of supported MT and LLM providers.
Prerequisites: Provider Credentials
Before creating any Profile, you need to store provider credentials in Smartling. There are two ways to obtain credentials:
-
Smartling-provisioned credentials - The Smartling team provides credentials for you (automatically available, no setup needed)
Smartling keys are available for a limited number of providers and may include usage restrictions, such as being intended for initial testing only. Refer to the Smartling documentation for the specific provider to confirm whether a Smartling key is available. If you have questions about Smartling keys, contact your Customer Success Manager or another Smartling representative.
- Bring Your Own Key (BYOK) - Add your own provider credentials to the Credentials page in the AI Hub
Learn more in MT and LLM Provider Credentials.
Smartling Auto Select
Your Smartling account comes with a pre-configured MT Profile called Smartling Auto Select. This Profile automatically selects the highest-performing MT provider for each language pair and requires no setup to get started.
Learn more about Smartling Auto Select.
How to access Profiles
To view, manage, or create Profiles:
- Click into the AI Hub from the top navigation bar of your Smartling dashboard.
- Click into the Profiles section.
From here, you can view and edit all existing MT and LLM Profiles and create new Profiles.
Next steps
Ready to create a new Profile?