POST /v1/embeddings — generate vector embeddings for text

The /v1/embeddings endpoint converts text into dense numerical vectors (embeddings) that capture semantic meaning. These vectors are useful for semantic search, retrieval-augmented generation (RAG), clustering, and similarity ranking. MonoRelay routes the request to the appropriate provider based on your model routing configuration and returns the embeddings in the standard OpenAI format.

Method and path

POST /v1/embeddings

Authentication

Include your Bearer token in the Authorization header.

Authorization: Bearer <your-access-token>

Request body

string

required

The embedding model to use, such as text-embedding-3-small or text-embedding-ada-002. Accepts aliases and model@provider syntax.

string | string[]

required

The text to embed. Pass a single string for one embedding, or an array of strings to embed multiple texts in a single request. All strings are processed as a batch by the upstream provider.

string

default:"float"

The format of the returned embeddings. Use "float" for a list of floating-point numbers, or "base64" for a base64-encoded binary representation. Not all providers support "base64".

Response

string

Always "list".

object[]

Array of embedding objects, one per input string, in the same order as the input.

Show data item properties

string

Always "embedding".

integer

Zero-based index of this embedding in the input array.

number[]

The embedding vector as an array of floating-point numbers. The dimensionality depends on the model.

string

The model name as returned by the upstream provider.

object

Token usage for the request.

Show usage properties

integer

Number of tokens in the input text.

integer

Total tokens processed (same as prompt_tokens for embeddings).

Examples

from openai import OpenAI

client = OpenAI(
    base_url="https://<host>/v1",
    api_key="<your-access-token>",
)

response = client.embeddings.create(
    model="text-embedding-3-small",
    input=["MonoRelay is an LLM relay server.", "It supports multiple providers."],
)

for item in response.data:
    print(f"Index {item.index}: {len(item.embedding)}-dimensional vector")

curl https://<host>/v1/embeddings \
  -H "Authorization: Bearer <token>" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "text-embedding-3-small",
    "input": "MonoRelay is an LLM relay server."
  }'

Error responses

Errors are returned as JSON with HTTP 503. The message field is prefixed with the provider name for easier debugging.

{
  "error": {
    "message": "[openai] No available keys for provider 'openai'",
    "type": "no_keys"
  }
}

Auto-generate your docs

Overview

OpenAI-Compatible

Anthropic-Compatible

Management API

POST /v1/embeddings — generate vector embeddings for text

Method and path

Authentication

Request body

Response

Examples

Error responses

​Method and path

​Authentication

​Request body

​Response

​Examples

​Error responses

Method and path

Authentication

Request body

Response

Examples

Error responses