Skip to main content

Configure and Select GenAI Models for Use with Rowan

Set desired models in Settings and change your selectin as you work. Rowan supports using a number of models with our GenAI features.

Laura Berwick avatar
Written by Laura Berwick
Updated today

Note that our GenAI features are only available on request, and some options may be unavailable at the direction of your organization. Contact your Rowan Patents Account Manage or Rowan Patents Help to gain access, ask questions, and discuss additional needs.

  1. Open Settings > Generative AI from the Tools menu or the Tools window


    OR

    Within the sidebar or chat, press the Settings gear icon to reach the GenAI Chat and Assistant controls.



    Note that each of the dropdowns provide the same options and behaviors as those illustrated below for the Tools > Settings > Generative AI tab.

  2. If desired, click the +Portkey and +Azure buttons to add one or more custom configurations of those types.


  3. Provide custom configuration details using the fields provided.

    Note that the Configuration Name parameter is a local name you create to identify the model and not a parameter sent externally.

  4. To delete a custom configuration, click the Delete button for that configuration



  5. Select your desired default models or model configurations for GenAI Chat and GenAI Assistants from the dropdowns provided.



    Note that your GenAI Chat and GenAI Assistant model selections can be updated as you work.​

    As of version 3.30.2, ​OpenAI GPT is accessed using a secure Clarivate OpenAI deployment. Per OpenAI policies (https://openai.com/policies/api-data-usage-policies), your information will NOT used to train the model and will not be stored by Clarivate's deployment beyond 30 days. Note that clicking Send is the only way your data will ever be sent for cloud AI processing. We will never collect and send your IP without your express request.

    Rowan's secure local language model (Llama 3.2, https://www.llama.com/docs/model-cards-and-prompt-formats/llama3_2/) is available to all users with access to our GenAI features. If the model fails to initialize, make sure the address to download the model (https://rowan-public-data.s3.us-east-1.amazonaws.com/language-models) has been whitelisted by your IT department, then restart the software.

    Contact your Account Manager with questions or to explore accessing your company's own GenAI model(s) in Rowan.

Did this answer your question?