diff --git a/docs/getting-started.md b/docs/getting-started.md
index c9eaa369cd130b2b5bfcf7efdb91980735b35f2c..5c21f8b779e31125de6051a639033acad5726b0c 100644
--- a/docs/getting-started.md
+++ b/docs/getting-started.md
@@ -20,6 +20,22 @@ or in a Jupyter notebook:
 !pip install tgclients
 ```
 
+## Databinding for TextGrid metadata
+
+*tgclients* provide a databinding generated with [xsdata](https://pypi.org/project/xsdata/) for dealing
+with [TextGrid metadata](https://doc.textgrid.de/Metadata/). This means that for *tgsearch* and *tgcrud*
+there are an implementation returning metadata with databinding (TextgridSearch, TextgridCrud) and
+one low level API without databinding (TextgridSearchRequest, TextgridCrudRequest).
+
+With the databinding you can navigate the XML metadata as Python object and do not need to use XML parsers
+or XPath. E.g: `result.object_value.generic.provided.title[0]` gives you access to the first title.
+
+To understande which metadata fields to find in what place of the TextGrid
+metadata model check the [cheat sheet](https://doc.textgrid.de/attachments/metadata/cheatsheet.pdf):
+
+[![TextGrid metadata cheat sheet](https://doc.textgrid.de/attachments/metadata/cheatsheet.png)](https://doc.textgrid.de/attachments/metadata/cheatsheet.png)
+
+
 ## Example Jupyter notebooks
 
 Here are some Jupyter Notebooks with some examples how to work with the tgclients.