Embed a document longer than your input limit

My embedding call worked fine in every tutorial example, then started failing the moment I pointed it at one of our longest real documents, and I had no idea why.

IntermediateRead 6 minDoing: an afternoon7 steps

▶ Start

What you'll have at the end

A function that embeds a document of any length without truncating or crashing.

Before you start

You need

A working call to an embedding model that already returns a vector for a short piece of text, plus a real document of yours long enough that you suspect it crosses that model's input limit.

Not covered

Deciding how to store each chunk's vector separately for passage-level search across a full retrieval index; this only covers building one function that returns a single, reliable vector for a document of any length.

Leans on

7 steps

Where to go next

Split your documents into chunks that help RAG give better answers

go there instead if you need every chunk to stay separately searchable, rather than merged into one document vector

Checked 15 Aug 2026

Part of the Embeddings cookbook