Skip to content


At First Derm we are always improving our deep learning models. In our new API we provide versioned models with the following name:


Where N is the version number of the model and R is the revision. For example autoderm_v1_0 is the version 1 of our model, autoderm_v1_1 is the same version but revised, and autoderm_v1_2 would be the second revision.

A revision means:

  • We are keeping the same model overall.
  • We may have removed certain classes or done some performance optimizations. But the metrics for the other classes have not changed.

A version means a new model. That can mean:

  • Adding new classes.
  • Using new neural network architectures.
  • Metrics have changed.

The most important thing to take into account is that when we add a new class, the model will have a new version.

How to use the different versions

You can simply change the URL parameter model to select which model you want to use for each API call. The default model is autoderm_v2_0.

The following URL paths will select different model versions:

  • \v1\query?model=autoderm_v2_0
  • \v1\query?model=autoderm_v2_1

You can also have a look at the examples section

You can select a different version at any time and for each image sent.


  • autoderm_v2_1
  • Added new diseases.
  • autoderm_v2_0
  • autoderm_v1_1
  • autoderm_v1_0

If you need more details about the models and are interested, you can contact us.