There are different ways to achieve a “contains” query.
1) You could use wildcard matching Lucene query syntax, as mentioned in this doc, however, this is potentially very slow for large indexes.
2) You could use a custom analyzer such as NGram or EdgeNGram. This will index all the permutations of the field which creates a larger index but makes queries much faster. Check out this blog article for more information: Custom analyzers in Azure Search
Note that the blog uses EdgeNGram which will give you a “prefix” query and not a “contains” query. Change it to “ngram” or “ngram_v2” instead. See the list of tokenizers available here: Add custom analyzers to string fields