Create Embeddings

POST /v1/embeddings Models: Embeddings View .md

Embed text in OpenRouter's embeddings shape, which is OpenAI's with input_type and an id added.

Batches: Up to 2048 inputs per request. The API splits them into batches the model's engine accepts and returns the embeddings in input order.

Queries and documents: Instruction-aware models such as qwen3-embedding-4b retrieve better when queries carry an instruction. Set input_type to search_query on queries and the model's instruction is prepended; documents (search_document, or no input_type) are sent unchanged. A query that already starts with the instruction's marker (e.g. Instruct:) is not prefixed again. The prefix's tokens are billed.

Long inputs: An input longer than the model's maximum input length is truncated to it, not rejected.

Body

application/json

OpenRouter's embeddings request. OpenRouter's provider, session_id and trace and OpenAI's user are accepted and ignored.

model string required

The embeddings model to use.
input string or array of strings or array of integers or array of array of integers required

A string, a list of strings, a list of token ids, or a list of token id lists. Up to 2048 inputs. Multimodal inputs are rejected.
One of 4
- string
- array of strings
- array of integers
- array of array of integers
encoding_format string

base64 returns each embedding as base64-encoded little-endian float32 values.

default floatfloatbase64
dimensions integer

Truncate embeddings to this many dimensions, for models trained to support it (e.g. Qwen3-Embedding).

at least 1
input_type string

search_query (or query) marks the inputs as search queries, so instruction-aware models get their query instruction. Any other value, or none, sends the inputs unchanged.

Responses

application/json · embeddings-response-payload

5 fields

id string required

Identifier for this response, prefixed emb-.
object string required

Always list.
data array of objects required

One entry per input, in input order.
3 fields
- object string required
  
  Always embedding.
- index integer required
  
  Position of the input in the request.
- embedding array of numbers or string required
  
  A list of floats, or a base64 string when encoding_format is base64.
  One of 2
  - array of numbers
  - string
model string required
usage object required
2 fields
- prompt_tokens integer
  
  Tokens the model processed, including any query instruction. This is the billed count.
- total_tokens integer

curl https://api.vultrinference.com/v1/embeddings \
  -H "Authorization: Bearer $VULTR_INFERENCE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen3-embedding-4b",
    "input": [
      "What is the capital of France?"
    ],
    "input_type": "search_query"
  }'

import os
import requests

response = requests.post(
    "https://api.vultrinference.com/v1/embeddings",
    headers={"Authorization": f"Bearer {os.environ['VULTR_INFERENCE_API_KEY']}"},
    json={
        "model": "qwen3-embedding-4b",
        "input": [
            "What is the capital of France?",
        ],
        "input_type": "search_query",
    },
)
print(response.json())

const response = await fetch("https://api.vultrinference.com/v1/embeddings", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.VULTR_INFERENCE_API_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    "model": "qwen3-embedding-4b",
    "input": [
      "What is the capital of France?"
    ],
    "input_type": "search_query"
  }),
});
console.log(await response.json());

Response · 200

{
    "id": "string",
    "object": "string",
    "data": [
        {
            "object": "string",
            "index": 0,
            "embedding": [
                null
            ]
        }
    ],
    "model": "string",
    "usage": {
        "prompt_tokens": 0,
        "total_tokens": 0
    }
}