AI functions for vectors, schema, and tokens

Intermediate

Use GetEmbedding, CosineSimilarity, NormalizeEmbedding, GetTableDDL, GetFieldsOnLayout, and GetTokenCount in AI workflows.

What you'll learn

Which functions create, convert, normalize, add, and subtract embeddings
Which functions expose layout fields and table schema
Why GetTokenCount is guidance, not a billing guarantee

FileMaker AI functions support both vector work and prompt preparation. Claris tests often ask what each function returns, especially CosineSimilarity, GetEmbedding, GetTableDDL, GetFieldsOnLayout, and GetTokenCount.

Stuck is a valid status

Need a second brain on this one?

If this lesson just collided with your real schema, script stack, or deadline, book consulting and turn the confusion into a concrete plan.

Book consulting

1/3

Use GetEmbedding for expression-level vectors

GetEmbedding sends input data to an embedding model and returns container data. It is useful when you want to store a query vector, compare vectors in a calculation, or prepare Query by Vector data.

FileMaker Script

Set Field [ Global::Query_Vector ;
  GetEmbedding ( "main-ai" ; "text-embedding-3-small" ; Global::Search_Text )
]

Embeddings and Perform Semantic Find RAG accounts and Perform RAG Action