Embeddings API

An embedding is a vector representation of a piece of data. Text embeddings are often used to capture the semantic meaning of a piece of text, where the distance between text embedding vectors is used to measure their relatedness.

Create Embeddings

Endpoint to create an embedding vector representation of a text input.

POST https://api.relax.ai/v1/embeddings

Example Request

curl

curl https://api.relax.ai/v1/embeddings \
  -H "Authorization: Bearer $RELAX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "input": "The capital city of the UK is London",
    "model": "Mistral-7b-embedding",
    "encoding_format": "float"
  }'

python

embed.py

import openai

client = OpenAI(
  api_key = RELAX_API_KEY,
  base_url = 'https://api.relax.ai/v1/',
)

response = client.embeddings.create(
    model="Mistral-7b-embedding",
    input="The capital city of the UK is London",
    encoding_format="float"
)

print(response)

node.js

embed.js

import OpenAI from "openai";

const openai = new OpenAI({
    apiKey: RELAX_API_KEY,
    baseURL: 'https://api.relax.ai/v1/'
  });

async function main() {
  const embedding = await openai.embeddings.create({
    model: "Mistral-7b-embedding",
    input: "The capital city of the UK is London",
    encoding_format: "float",
  });

  console.log(embedding);
}

main();

---

Response

Returns an embedding object, which contains the embedding vector of the text input.

Embedding Response

{
 "object": "list",
 "data": [
  {
   "object": "embedding",
   "embedding": [
    -0.002363205,
    0.005371202,
    ...
   ],
   "index": 0
  }
 ],
 "model": "Mistral-7b-embedding",
 "usage": {
  "prompt_tokens": 10,
  "completion_tokens": 0,
  "total_tokens": 10,
  "prompt_tokens_details": null,
  "completion_tokens_details": null
 }
}

---

Request Body

The following parameters can be included in the request body:

---

Create Embeddings Request Body

model

• Type: string
• Required: Yes
• Description: The model name to use for generating the completion.

---

input

• Type: string or array
• Required: Yes
• Description: Input text to embed. To embed multiple inputs in a single request, pass an array of strings or array of token arrays. The input must not exceed the max token length of the model (4096 tokens for the Mistral-7b-embedding model) and cannot be an empty string.

---

encoding_format

• Type: string
• Required: No
• Description: Embeddings vector format. Either float or base64.

---