Short answerChunking splits your documents into passages, and embeddings turn each passage into numbers that represent its meaning. Retrieval finds the passages closest to a question, so well-structured, self-contained sections work best.
Chunks
A chunk is a passage of your document, often a few paragraphs. Systems send only the best-matching chunks to the model. If an answer is split across two chunks or buried in a long table, it may be missed.
Embeddings
An embedding is a list of numbers that represents the meaning of a piece of text. Similar meanings give similar numbers. A question is embedded the same way and compared to the stored chunks in a vector store.
Write files that chunk well
- Use headings and short sections.
- Make each section self-contained: avoid "as above".
- Put facts in sentences, not only in images or complex tables.
- Repeat the key term (product name, policy name) in the section.
If answers are missing, see knowledge file problems. Practical guidance: splitting files.
Check your GPT before you move it
Paste your instructions, file names and Actions schema into the free checker. It runs in your browser and builds a portable bundle. Nothing is uploaded.
Open the free checker