Create an agent

Create an agent with metadata, runtime configuration, and an optional file tree. Creation is atomic: either the agent is fully created or no state is persisted.

POST /agents

Create an agent

cURL

curl --request POST \
  --url https://api.smith.langchain.com/v1/deepagents/agents \
  --header 'Content-Type: application/json' \
  --header 'X-Api-Key: <api-key>' \
  --data '
{
  "name": "research-assistant",
  "description": "Research assistant that can search the web and summarize sources.",
  "permissions": {
    "identity": "personal",
    "visibility": "user"
  },
  "runtime": {
    "model": {
      "model_id": "openai:gpt-5.5"
    }
  },
  "instructions": "You are a careful research assistant. Search for sources, keep notes, and return concise answers with citations.",
  "tools": {
    "tools": [
      {
        "name": "tavily_web_search",
        "mcp_server_url": "https://tools.langchain.com",
        "mcp_server_name": "Fleet",
        "display_name": "tavily_web_search"
      }
    ],
    "interrupt_config": {
      "https://tools.langchain.com::tavily_web_search::Fleet": true
    }
  }
}
'

import requests

url = "https://api.smith.langchain.com/v1/deepagents/agents"

payload = {
    "name": "research-assistant",
    "description": "Research assistant that can search the web and summarize sources.",
    "permissions": {
        "identity": "personal",
        "visibility": "user"
    },
    "runtime": { "model": { "model_id": "openai:gpt-5.5" } },
    "instructions": "You are a careful research assistant. Search for sources, keep notes, and return concise answers with citations.",
    "tools": {
        "tools": [
            {
                "name": "tavily_web_search",
                "mcp_server_url": "https://tools.langchain.com",
                "mcp_server_name": "Fleet",
                "display_name": "tavily_web_search"
            }
        ],
        "interrupt_config": { "https://tools.langchain.com::tavily_web_search::Fleet": True }
    }
}
headers = {
    "X-Api-Key": "<api-key>",
    "Content-Type": "application/json"
}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

const options = {
  method: 'POST',
  headers: {'X-Api-Key': '<api-key>', 'Content-Type': 'application/json'},
  body: JSON.stringify({
    name: 'research-assistant',
    description: 'Research assistant that can search the web and summarize sources.',
    permissions: {identity: 'personal', visibility: 'user'},
    runtime: {model: {model_id: 'openai:gpt-5.5'}},
    instructions: 'You are a careful research assistant. Search for sources, keep notes, and return concise answers with citations.',
    tools: {
      tools: [
        {
          name: 'tavily_web_search',
          mcp_server_url: 'https://tools.langchain.com',
          mcp_server_name: 'Fleet',
          display_name: 'tavily_web_search'
        }
      ],
      interrupt_config: {'https://tools.langchain.com::tavily_web_search::Fleet': true}
    }
  })
};

fetch('https://api.smith.langchain.com/v1/deepagents/agents', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

<?php

$curl = curl_init();

curl_setopt_array($curl, [
  CURLOPT_URL => "https://api.smith.langchain.com/v1/deepagents/agents",
  CURLOPT_RETURNTRANSFER => true,
  CURLOPT_ENCODING => "",
  CURLOPT_MAXREDIRS => 10,
  CURLOPT_TIMEOUT => 30,
  CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
  CURLOPT_CUSTOMREQUEST => "POST",
  CURLOPT_POSTFIELDS => json_encode([
    'name' => 'research-assistant',
    'description' => 'Research assistant that can search the web and summarize sources.',
    'permissions' => [
        'identity' => 'personal',
        'visibility' => 'user'
    ],
    'runtime' => [
        'model' => [
                'model_id' => 'openai:gpt-5.5'
        ]
    ],
    'instructions' => 'You are a careful research assistant. Search for sources, keep notes, and return concise answers with citations.',
    'tools' => [
        'tools' => [
                [
                                'name' => 'tavily_web_search',
                                'mcp_server_url' => 'https://tools.langchain.com',
                                'mcp_server_name' => 'Fleet',
                                'display_name' => 'tavily_web_search'
                ]
        ],
        'interrupt_config' => [
                'https://tools.langchain.com::tavily_web_search::Fleet' => true
        ]
    ]
  ]),
  CURLOPT_HTTPHEADER => [
    "Content-Type: application/json",
    "X-Api-Key: <api-key>"
  ],
]);

$response = curl_exec($curl);
$err = curl_error($curl);

curl_close($curl);

if ($err) {
  echo "cURL Error #:" . $err;
} else {
  echo $response;
}

package main

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

func main() {

	url := "https://api.smith.langchain.com/v1/deepagents/agents"

	payload := strings.NewReader("{\n  \"name\": \"research-assistant\",\n  \"description\": \"Research assistant that can search the web and summarize sources.\",\n  \"permissions\": {\n    \"identity\": \"personal\",\n    \"visibility\": \"user\"\n  },\n  \"runtime\": {\n    \"model\": {\n      \"model_id\": \"openai:gpt-5.5\"\n    }\n  },\n  \"instructions\": \"You are a careful research assistant. Search for sources, keep notes, and return concise answers with citations.\",\n  \"tools\": {\n    \"tools\": [\n      {\n        \"name\": \"tavily_web_search\",\n        \"mcp_server_url\": \"https://tools.langchain.com\",\n        \"mcp_server_name\": \"Fleet\",\n        \"display_name\": \"tavily_web_search\"\n      }\n    ],\n    \"interrupt_config\": {\n      \"https://tools.langchain.com::tavily_web_search::Fleet\": true\n    }\n  }\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("X-Api-Key", "<api-key>")
	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))

}

HttpResponse<String> response = Unirest.post("https://api.smith.langchain.com/v1/deepagents/agents")
  .header("X-Api-Key", "<api-key>")
  .header("Content-Type", "application/json")
  .body("{\n  \"name\": \"research-assistant\",\n  \"description\": \"Research assistant that can search the web and summarize sources.\",\n  \"permissions\": {\n    \"identity\": \"personal\",\n    \"visibility\": \"user\"\n  },\n  \"runtime\": {\n    \"model\": {\n      \"model_id\": \"openai:gpt-5.5\"\n    }\n  },\n  \"instructions\": \"You are a careful research assistant. Search for sources, keep notes, and return concise answers with citations.\",\n  \"tools\": {\n    \"tools\": [\n      {\n        \"name\": \"tavily_web_search\",\n        \"mcp_server_url\": \"https://tools.langchain.com\",\n        \"mcp_server_name\": \"Fleet\",\n        \"display_name\": \"tavily_web_search\"\n      }\n    ],\n    \"interrupt_config\": {\n      \"https://tools.langchain.com::tavily_web_search::Fleet\": true\n    }\n  }\n}")
  .asString();

require 'uri'
require 'net/http'

url = URI("https://api.smith.langchain.com/v1/deepagents/agents")

http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true

request = Net::HTTP::Post.new(url)
request["X-Api-Key"] = '<api-key>'
request["Content-Type"] = 'application/json'
request.body = "{\n  \"name\": \"research-assistant\",\n  \"description\": \"Research assistant that can search the web and summarize sources.\",\n  \"permissions\": {\n    \"identity\": \"personal\",\n    \"visibility\": \"user\"\n  },\n  \"runtime\": {\n    \"model\": {\n      \"model_id\": \"openai:gpt-5.5\"\n    }\n  },\n  \"instructions\": \"You are a careful research assistant. Search for sources, keep notes, and return concise answers with citations.\",\n  \"tools\": {\n    \"tools\": [\n      {\n        \"name\": \"tavily_web_search\",\n        \"mcp_server_url\": \"https://tools.langchain.com\",\n        \"mcp_server_name\": \"Fleet\",\n        \"display_name\": \"tavily_web_search\"\n      }\n    ],\n    \"interrupt_config\": {\n      \"https://tools.langchain.com::tavily_web_search::Fleet\": true\n    }\n  }\n}"

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

201

basic

{ "id": "dfa71115-c971-4197-8f02-20826fbc6f40", "name": "research-assistant", "description": "Research assistant that can search the web and summarize sources.", "owner_id": "33499cf9-4c02-49af-909d-b8f42e051eab", "created_at": "2026-05-15T20:50:32.708302+00:00", "updated_at": "2026-05-15T20:50:32.708302+00:00", "permissions": { "identity": "personal", "visibility": "user", "tenant_access_level": "read" }, "runtime": { "model": { "model_id": "openai:gpt-5.5" } }, "revision": "13ac11f11da376e588cb84825ed88d05298f27aaaa90c6f694abf572f34161d9", "instructions": "You are a careful research assistant. Search for sources, keep notes, and return concise answers with citations.", "tools": { "tools": [ { "name": "tavily_web_search", "mcp_server_url": "https://tools.langchain.com", "mcp_server_name": "Fleet", "display_name": "tavily_web_search" } ], "interrupt_config": { "https://tools.langchain.com::tavily_web_search::Fleet": true } } }

Authorizations

X-Api-Key · string · header · default: LANGSMITH_API_KEY · required

Headers

X-User-Timezone · string

IANA timezone name (for example, America/Los_Angeles) stamped into the agent's runtime configuration when the agent is created or updated. The runtime uses this timezone to inject a localized current-date line into the agent's system prompt, so the agent reasons about dates in the user's local time. Optional. Defaults to UTC when omitted; invalid values fall back to UTC.

Example:

"America/Los_Angeles"

Query Parameters

include_files · boolean · default: false

Include the raw file map in the agent response. Accepted truthy values are true and 1.

Body

application/json

name · string · required

Human-readable agent name. Must not be blank.

description · string

Short agent description.

permissions · object

Agent visibility and sharing configuration. When omitted on create, defaults to personal identity, tenant visibility, and read tenant access.

Show child attributes

runtime · object

Show child attributes

instructions · string

Deprecated alias for system_prompt. Accepted for backwards compatibility; system_prompt takes precedence when both are set.

tools · object

Show child attributes

subagents · object[]

Subagents. Each entry is written to subagents/<name>/AGENTS.md and, when provided, subagents/<name>/tools.json.

Show child attributes

skills · object[]

Inline skills. Each entry is written to skills/<name>/SKILL.md and supporting files.

Show child attributes

extras · object

Caller-defined key-value metadata attached to the agent.

files · object

Raw file map for paths not covered by typed fields. Keys are relative file paths. Setting a typed field and the corresponding files entry returns 422.

Show child attributes

backend · object

Show child attributes

system_prompt · string

Agent system prompt. Written to AGENTS.md. Preferred over deprecated instructions when both are provided.

configurable · object

Caller-defined runtime configuration merged into the agent config. Typed runtime fields win on key collision.

Response

201

application/json

Agent created.

id · string<uuid>

name · string

description · string

owner_id · string<uuid>

created_at · string<date-time>

updated_at · string<date-time>

permissions · object

Agent visibility and sharing configuration. When omitted on create, defaults to personal identity, tenant visibility, and read tenant access.

Show child attributes

runtime · object

Show child attributes

extras · object

access_level · enum<string>

Caller access level for this agent.

Available options: READ, RUN, WRITE

is_owner · boolean

Whether the authenticated caller owns this agent.

backend · object

Show child attributes

revision · string

Revision token for the latest file tree commit.

instructions · string

Deprecated alias echoed alongside system_prompt for backwards compatibility.

tools · object

Show child attributes

subagents · object[]

Show child attributes

skills · object[]

Show child attributes

files · object

Raw file map for paths not covered by typed fields. Keys are relative file paths. Setting a typed field and the corresponding files entry returns 422.

Show child attributes

system_prompt · string

Agent system prompt parsed from AGENTS.md.