Search
Rank documents or passages by semantic similarity to a query.
AI & ML · Const
Embedding
Embeddings represent text as vectors that preserve aspects of content or meaning. They support similarity-based tasks such as search, clustering, recommendations, anomaly detection, and classification.
How it works
An embedding is a vector representation of text that preserves aspects of its content or meaning. Send one or more text inputs with an embedding model to the embeddings endpoint; the response contains a vector for each input, which can then be compared or indexed for similarity-based applications.
Syntax
client.embeddings.create(input=[text], model="text-embedding-3-small")Parameters
| Name | Type | Description |
|---|---|---|
input | string | array of strings | Text, or an array of text inputs, to embed. |
model | string | The identifier of the embedding model, such as text-embedding-3-small. |
Returns
An embeddings response containing a data array; each item includes the generated embedding vector.
Key points
- Embeddings convert text into vectors whose relative closeness represents semantic relatedness.
- Typical applications include semantic search, clustering, recommendations, anomaly detection, diversity measurement, and classification.
- Call the embeddings endpoint with the input text and model identifier, such as text-embedding-3-small.
- Embedding models accept text and return numerical vectors suitable for downstream similarity operations.
- Check the model index for the input limit of the chosen embedding model.
Common mistakes
- Keep each input below the selected embedding model’s maximum context length.
- The request must specify both the text input and an embedding model.
Example: Create a text embedding
from openai import OpenAI
client = OpenAI()
response = client.embeddings.create(
input=["The food was delicious"],
model="text-embedding-3-small"
)
print(len(response.data[0].embedding))Creates an embedding for one text string and prints the returned vector’s dimensionality.
When to use it
Clustering
Group text items by semantic similarity.
Recommendations
Recommend related items based on their text representations.
Classification
Assign text to the label with the most similar embedding.
Related
text-embedding-3-smallembeddings APIsemantic searchclusteringrecommendationsclassification