License Management


By default, the Annotation Lab allows access to community pre-trained models and embeddings. Those are available on the Models Hub page. To gain access to licensed resources (e.g. pre-trained models and embeddings) admin user can import a license (Healthcare, Finance, Legal, or Visual NLP) which will activate additional features:

  • Access to licensed models for pre-annotation
  • Access to healthcare, finance, and legal embeddings
  • Access to rules
  • Access to optimized annotators
  • Access to training custom models using licensed embeddings

The admin user can upload a Spark NLP license JSON file by visiting the License page. The license is generated by the John Snow Labs license server and is available on Once a valid license is uploaded, all the licensed (Healthcare, Finance, Legal, and Visual NLP) models and embeddings become available for download. The License page shows the history of license uploads with detailed information like License Info, Status, Renewal Date, and License Secrets.

Support for Floating Licenses

Annotation Lab supports floating licenses with different scopes (ocr: training, ocr: inference, healthcare: inference, healthcare: training, finance: inference, finance: training, legal: inference, legal: training). Depending on the scope of the available license, users can perform model training and/or deploy pre-annotation servers. Licenses are a must for training Healthcare, Finance, and Legal models and deploying these models as pre-annotation servers. Floating licenses can be acquired on self-service via

One floating license is bound to only one server (pre-annotation server, OCR server, training job) at a time. To run multiple model training jobs and/or pre-annotations servers, users must provide multiple floating licenses.

Annotation Lab supports either floating licenses or air-gapped licenses. Mixing floating and air-gapped licenses on the same Annotation Lab instance is not allowed.

Usage of NLP Licenses

The number of available floating licenses can influence the creation of multiple training and pre-annotation servers. For example, to deploy 5 pre-annotation servers using Spark NLP for Healthcare models or embeddings, across 5 different projects, you will need 5 floating licenses.

Since one floating license can only be used for one server, it is not possible to deploy a pre-annotation server and then trigger training from the same project when only one license is available. In this case, the pre-annotation server has to be deleted first, and then the training can be started.

Those restrictions do not apply when using Spark NLP models and embeddings.

Last updated