Create Index - Upstash Documentation

post/vector/index

Request example

curl --request POST \
  --url https://api.upstash.com/v2/vector/index \
  --header 'Authorization: Basic <encoded-value>' \
  --header 'Content-Type: application/json' \
  --data '{
  "name": "myindex",
  "region": "us-east-1",
  "similarity_function": "COSINE",
  "dimension_count": 1024,
  "type": "payg",
  "embedding_model": "BGE_M3",
  "index_type": "HYBRID",
  "sparse_embedding_model": "BM25"
}'

import requests

url = "https://api.upstash.com/v2/vector/index"
headers = {
    "Authorization": "Basic <encoded-value>",
    "Content-Type": "application/json",
}
payload = {
    "name": "myindex",
    "region": "us-east-1",
    "similarity_function": "COSINE",
    "dimension_count": 1024,
    "type": "payg",
    "embedding_model": "BGE_M3",
    "index_type": "HYBRID",
    "sparse_embedding_model": "BM25"
}

response = requests.request("POST", url, headers=headers, json=payload)

print(response.text)

const response = await fetch('https://api.upstash.com/v2/vector/index', {
  method: 'POST',
  headers: {
    'Authorization': 'Basic <encoded-value>',
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    "name": "myindex",
    "region": "us-east-1",
    "similarity_function": "COSINE",
    "dimension_count": 1024,
    "type": "payg",
    "embedding_model": "BGE_M3",
    "index_type": "HYBRID",
    "sparse_embedding_model": "BM25"
  })
});

const data = await response.json();
console.log(data);

<?php

$curl = curl_init();

curl_setopt_array($curl, [
  CURLOPT_URL => 'https://api.upstash.com/v2/vector/index',
  CURLOPT_RETURNTRANSFER => true,
  CURLOPT_CUSTOMREQUEST => 'POST',
  CURLOPT_POSTFIELDS => '{
  "name": "myindex",
  "region": "us-east-1",
  "similarity_function": "COSINE",
  "dimension_count": 1024,
  "type": "payg",
  "embedding_model": "BGE_M3",
  "index_type": "HYBRID",
  "sparse_embedding_model": "BM25"
}',
  CURLOPT_HTTPHEADER => [
    'Authorization: Basic <encoded-value>',
    'Content-Type: application/json',
  ],
]);

$response = curl_exec($curl);
curl_close($curl);

echo $response;

package main

import (
	"fmt"
	"io"
	"net/http"
	"strings"
)

func main() {
	url := "https://api.upstash.com/v2/vector/index"
	payload := strings.NewReader("{\n  \"name\": \"myindex\",\n  \"region\": \"us-east-1\",\n  \"similarity_function\": \"COSINE\",\n  \"dimension_count\": 1024,\n  \"type\": \"payg\",\n  \"embedding_model\": \"BGE_M3\",\n  \"index_type\": \"HYBRID\",\n  \"sparse_embedding_model\": \"BM25\"\n}")

	req, _ := http.NewRequest("POST", url, payload)
	req.Header.Add("Authorization", "Basic <encoded-value>")
	req.Header.Add("Content-Type", "application/json")

	res, _ := http.DefaultClient.Do(req)
	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))
}

HttpRequest request = HttpRequest.newBuilder()
    .uri(URI.create("https://api.upstash.com/v2/vector/index"))
    .header("Authorization", "Basic <encoded-value>")
    .header("Content-Type", "application/json")
    .method("POST", HttpRequest.BodyPublishers.ofString("{\n  \"name\": \"myindex\",\n  \"region\": \"us-east-1\",\n  \"similarity_function\": \"COSINE\",\n  \"dimension_count\": 1024,\n  \"type\": \"payg\",\n  \"embedding_model\": \"BGE_M3\",\n  \"index_type\": \"HYBRID\",\n  \"sparse_embedding_model\": \"BM25\"\n}"))
    .build();

HttpResponse<String> response = HttpClient.newHttpClient()
    .send(request, HttpResponse.BodyHandlers.ofString());

System.out.println(response.body());

require 'uri'
require 'net/http'

url = URI('https://api.upstash.com/v2/vector/index')

http = Net::HTTP.new(url.host, url.port)
http.use_ssl = url.scheme == 'https'

request = Net::HTTP::Post.new(url)
request['Authorization'] = 'Basic <encoded-value>'
request['Content-Type'] = 'application/json'
request.body = '{
  "name": "myindex",
  "region": "us-east-1",
  "similarity_function": "COSINE",
  "dimension_count": 1024,
  "type": "payg",
  "embedding_model": "BGE_M3",
  "index_type": "HYBRID",
  "sparse_embedding_model": "BM25"
}'

response = http.request(request)
puts response.read_body

Response

{
  "customer_id": "example@upstash.com",
  "id": "0639864f-ece6-429c-8118-86a287b0e808",
  "name": "myindex",
  "similarity_function": "COSINE",
  "dimension_count": 384,
  "embedding_model": "BGE_SMALL_EN_V1_5",
  "sparse_embedding_model": "BM25",
  "endpoint": "glowing-baboon-15797-us1",
  "token": "QkZGAsWp2tdW0tdC0zNzM1LWV1MkFkNQzB1ExUb3hOekF0TVRJbFpMDNLVSm1GZw==",
  "read_only_token": "QkZGRk1heGSKC0MtdRlZC0zNzM1LWTj3pAV0Wm1aZ01p05qY3RNR0U0TkRtRt2s9azJU",
  "type": "fixed",
  "region": "us-east-1",
  "max_vector_count": 5210000,
  "max_daily_updates": 1000000,
  "max_daily_queries": 1000000,
  "max_monthly_bandwidth": -1,
  "max_writes_per_second": 1000,
  "max_query_per_second": 1000,
  "max_reads_per_request": 1000,
  "max_writes_per_request": 1000,
  "max_total_metadata_size": 53687091200,
  "reserved_price": 60,
  "creation_time": 1753207106,
  "index_type": "DENSE",
  "throughput_vector": [
    {
      "x": "2025-09-04 14:55:00.000 +0000 UTC",
      "y": 0
    },
    {
      "x": "2025-09-04 14:56:00.000 +0000 UTC",
      "y": 0
    }
  ]
}

post/vector/index

Authorization

Authorizationstringheaderrequired#

Basic authentication header of the form Basic <encoded-value>, where <encoded-value> is the base64-encoded username:password.

Body

application/json

namestringrequired#

Name of the index

Example:"myindex"

regionenum<string>required#

Region of the database

Options:eu-west-1us-east-1us-central1

Example:"us-east-1"

similarity_functionenum<string>required#

Similarity function that's used to calculate the distance between two vectors

Options:COSINEEUCLIDEANDOT_PRODUCT

Example:"COSINE"

dimension_countnumberrequired#

The amount of values in a single vector

Example:1024

typeenum<string>#

The payment plan of your index

Options:paygfixedpaid

Example:"payg"

embedding_modelenum<string>#

The embedding model to use for the index

Options:BGE_SMALL_EN_V1_5BGE_BASE_EN_V1_5BGE_LARGE_EN_V1_5BGE_M3

Example:"BGE_M3"

index_typeenum<string>#

The type of the vector index

Options:DENSESPARSEHYBRID

Example:"HYBRID"

sparse_embedding_modelenum<string>#

The sparse embedding model to be used for indexes

Options:BM25BGE_M3

Example:"BM25"

Responses

application/json

200

Index created successfully

customer_idstring#

The associated ID of the owner of the index

Example:"example@upstash.com"

idstring#

Unique ID of the index

Example:"0639864f-ece6-429c-8118-86a287b0e808"

namestring#

The name of the index

Example:"myindex"

similarity_functionenum<string>#

Similarity function that's used to calculate the distance between two vectors

Options:COSINEEUCLIDEANDOT_PRODUCT

Example:"COSINE"

dimension_countnumber#

The amount of values in a single vector

Example:384

embedding_modelenum<string>#

The predefined embedding model to vectorize your plain text

Options:BGE_SMALL_EN_V1_5BGE_BASE_EN_V1_5BGE_LARGE_EN_V1_5BGE_M3

Example:"BGE_SMALL_EN_V1_5"

sparse_embedding_modelenum<string>#

The sparse embedding model to be used for indexes

Options:BM25BGE_M3

Example:"BM25"

endpointstring#

The REST endpoint of the index

Example:"glowing-baboon-15797-us1"

tokenstring#

The REST authentication token for the index

Example:"QkZGAsWp2tdW0tdC0zNzM1LWV1MkFkNQzB1ExUb3hOekF0TVRJbFpMDNLVSm1GZw=="

read_only_tokenstring#

The REST authentication read only token for the index

Example:"QkZGRk1heGSKC0MtdRlZC0zNzM1LWTj3pAV0Wm1aZ01p05qY3RNR0U0TkRtRt2s9azJU"

typeenum<string>#

The payment plan of the index

Options:freepaygfixed

Example:"fixed"

regionenum<string>#

The region where the index is currently deployed

Options:eu-west-1us-east-1us-central1

Example:"us-east-1"

max_vector_countnumber#

The number of maximum that your index can contain

Example:5210000

max_daily_updatesnumber#

The number of maximum update operations you can perform in a day. Only upsert operations are included in update count.

Example:1000000

max_daily_queriesnumber#

The number of maximum query operations you can perform in a day. Only query operations are included in query count.

Example:1000000

max_monthly_bandwidthnumber#

The maximum amount of monthly bandwidth for the index. Unit is bytes. -1 if the limit is unlimited.

Example:-1

max_writes_per_secondnumber#

The number of maximum write operations you can perform per second. Only upsert operations are included in write count.

Example:1000

max_query_per_secondnumber#

The number of maximum query operations you can perform per second. Only query operations are included in query count.

Example:1000

max_reads_per_requestnumber#

The number of maximum vectors in a read operation. Query and fetch operations are included in read operations.

Example:1000

max_writes_per_requestnumber#

The number of maximum vectors in a write operation. Only upsert operations are included in write operations.

Example:1000

max_total_metadata_sizenumber#

The amount of maximum size for the total metadata sizes in your index.

Example:53687091200

reserved_pricenumber#

Monthly pricing of your index. Only available for fixed and pro plans.

Example:60

creation_timenumber#

The creation time of the vector index in UTC as unix timestamp.

Example:1753207106

index_typeenum<string>#

The type of the vector index

Options:DENSESPARSEHYBRID

Example:"DENSE"

throughput_vectorobject[]#

Throughput data for the vector index over time

Show child attributes

xstringrequired#

Timestamp when measurement was taken

Example:"2023-05-22 10:59:23.426 +0000 UTC"

ynumberrequired#

The measured value

Example:320

Example:

[  {    "x": "2025-09-04 14:55:00.000 +0000 UTC",    "y": 0  },  {    "x": "2025-09-04 14:56:00.000 +0000 UTC",    "y": 0  }]

Get Vector Stats Rename Index