✨ Vertex is now free! Learn more here
Docs
Endpoints
Search Profiles

Search Profiles

Search Profiles provides a list of pubkeys matching the provided search term. This service powers the search of npub.world .

Example use cases:

  • Searching profiles
  • Auto-completion when tagging someone

Transport

This service operates exclusively using signed Nostr Events for both requests and responses. The two communication flows differ only in the transport mechanism used to send and receive these events.

Transport Endpoint Send Receive
Nostr (websocket) wss://relay.vertexlab.io Publish the request event Fetch the response event with a filter
HTTP https://relay.vertexlab.io/api/v1/dvms POST with the request event in the body parse the response event from the response’s body

Example flow using Nostr 

# Publish the request
nak event \
  -k 5315 \
  --tag param="search;jack" \
  --tag param="limit=10" \
  wss://relay.vertexlab.io
# Fetch the response event by the request ID
nak req -k 6315 -k 7000 --tag e=<request_ID> wss://relay.vertexlab.io

Example flow using the API 

# Generate the request
REQUEST=$(nak event \
  -k 5315 \
  --tag param="search;jack" \
  --tag param="limit=10"
)

# POST the request
curl -X POST "https://relay.vertexlab.io/api/v1/dvms" \
     -H "Content-Type: application/json" \
     -d "$REQUEST"

Kinds

  • Request: 5315
  • Response: 6315
  • Error: 7000

Request

Parameters

All parameters must be passed as strings in the tags of the Request event, with the following format:

["param", "<name>", "<value>"]
Param Description Default Value Max
search (required) The search terms - one search
sort Algorithm used to sort results "globalPagerank" one sort
source Pubkey that provides the “point of view” for personalized algorithms The pubkey signing the DVM request one source
limit Maximum number of results returned "5" "100"

The search must be longer than three characters.
Sorting algorithms can be found here.

Cost

The cost is measured in Vertex credits and it solely depends on the sort parameter. For the $ prices, check out our pricing.

Sort Credits
followerCount 1
globalPagerank 1
personalizedPagerank 10

Example request 

{
  "id": "7a0058...cff3e0",
  "pubkey": "79be66...f81798",
  "created_at": 1740754324,
  "kind": 5315,
  "tags": [
    [
      "param",
      "search",
      "jack",
    ],
  ],
  "content": "",
  "sig": "e64ce6...b25bd9"
}

Response

Tags

Tag Description
e The event ID of the request
p The pubkey that signed the request
sort The sorting algorithm specified in the request
source The source specified in the request (present only if sort="personalizedPagerank")
nodes The number of nodes in the graph at the time the request was made

Content

The content field is a JSON-stringified array of objects, each formatted as:

Field Type Description
pubkey string a nostr hex pubkey
rank float the rank computed with the sort algorithm

Pubkeys are sorted in descending order by their rank.
The rank is a combination of two components:

  • search_rank: measures how well the search term matches the profile.
  • reputation_rank: determined by the sort parameter.

The combined rank is calculated with the formula:

rank = |search_rank|³ × reputation_rank

To learn more you can check out our code .

Example response 

{
  "id": "6fa475...a0f9ed",
  "pubkey": "5fc48a...ccaeb4",
  "created_at": 1740754324,
  "kind": 6315,
  "tags": [
    [
      "e",
      "7a0058...cff3e0"
    ],
    [
      "p",
      "79be66...f81798"
    ],
    [
      "sort",
      "globalPagerank"
    ],
    [
      "nodes",
      "317238"
    ],
  ],
  "content":"[{\"pubkey\":\"82341f882b6eabcd2ba7f1ef90aad961cf074af15b9ef44a09f9d2a8fbfbe6a2\",\"rank\":8.224848486131593},{\"pubkey\":\"c4eabae1be3cf657bc1855ee05e69de9f059cb7a059227168b80b89761cbc4e0\",\"rank\":2.120725804894317},{\"pubkey\":\"a1fc5dfd7ffcf563c89155b466751b580d115e136e2f8c90e8913385bbedb1cf\",\"rank\":0.2690585544819},{\"pubkey\":\"c5fb6ecc876e0458e3eca9918e370cbcd376901c58460512fe537a46e58c38bb\",\"rank\":0.14113029299502666},{\"pubkey\":\"5e5fc1434c928bcdcba6f801859d5238341093291980fd36e33b7416393d5a2c\",\"rank\":0.12754991708985827}]",
  "sig": "bc856a...898458"
}

Formatted content JSON:

[
    {
      "pubkey": "82341f...fbe6a2",
      "rank": 8.224848486131593
    },
    {
      "pubkey": "c4eaba...cbc4e0",
      "rank": 2.120725804894317
    },
    {
      "pubkey": "a1fc5d...edb1cf",
      "rank": 0.2690585544819
    },
    {
      "pubkey": "c5fb6e...8c38bb",
      "rank": 0.14113029299502666
    },
    {
      "pubkey": "5e5fc1...3d5a2c",
      "rank": 0.12754991708985827
    },
]

Error

Tags

Tag Description
e The event ID of the request
p The pubkey that signed the request
status The error type and error message

Example error 

{
  "kind": 7000,
  "tags": [
      [
        "e",
        "1cd2c7...5f7f49"
      ],
      [
        "p",
        "79be66...f81798"
      ],
      [
        "status",
        "error",
        "the search term must be longer than three characters"
      ],
  ],
  // ...
}

Example code 

package main

import (
	"context"
	"fmt"

	"github.com/nbd-wtf/go-nostr"
)

func main() {
	// step 1: connect to vertex
	vertex := "wss://relay.vertexlab.io"
	relay, err := nostr.RelayConnect(context.Background(), vertex)
	if err != nil {
		panic(err)
	}

	// step 2: generate the request and sign it with your vertex key.
	// Be careful to store your key securely.
	searchProfiles := &nostr.Event{
		Kind:      5315,
		CreatedAt: nostr.Now(),
		Tags: nostr.Tags{
			{"param", "search", "pip"},
			{"param", "limit", "7"},
		},
	}

	err = searchProfiles.Sign("YOUR_SECRET_KEY")
	if err != nil {
		panic(err)
	}

	// step 3: publish the request to the vertex relay
	err = relay.Publish(ctx, *searchProfiles)
	if err != nil {
		panic(err)
	}

	// step 3: fetch the response
	filter := nostr.Filter{
		Kinds: []int{6315, 7000},
		Tags: nostr.TagMap{
			"e": {searchProfiles.ID},
		},
	}

	responses, err := relay.QueryEvents(ctx, filter)
	if err != nil {
		panic(err)
	}

	// extract the first response
	response := <-responses

	// step 4: use the response in your app
	fmt.Printf("response: %v\n", response)
}
import { Relay, finalizeEvent } from 'nostr-tools';

try {
  // step 1: connect to vertex
  const vertex = 'wss://relay.vertexlab.io';
  const relay = new Relay(vertex);
  await relay.connect();

  // step 2: generate the request and sign it with your vertex key.
  let searchProfiles = {
    kind: 5315,
    created_at: Math.floor(Date.now() / 1000),
    tags: [
      ['param', 'target', 'npub176p7sup477k5738qhxx0hk2n0cty2k5je5uvalzvkvwmw4tltmeqw7vgup'],
      ['param', 'limit', '7'],
    ],
    content: '',
  };

  searchProfiles = finalizeEvent(searchProfiles, 'YOUR_SECRET_KEY')

  // Step 3: publish the request
  await relay.publish(searchProfiles);

  // Step 4: fetch the response
  const filter = {
    kinds: [6315, 7000],
    '#e': [searchProfiles.id],
  };

  const sub = relay.subscribe([filter], {
      onevent(response) {
        // step 5: use it in your app
          console.log('response:', response);
      },
      oneose() {
        resolve(events);
        sub.close();
      },
  })

} catch(err) {
  console.log('error:', err)
}
Rank Profiles