***

id: d4233f2c-adfc-4eed-bbc2-62bd6619e91e
title: Record
sidebar-title: Record
slug: /cxml/reference/voice/record
position: 1
max-toc-depth: 3
----------------

The `<Record>` verb creates an audio file with the caller's voice and returns the URL to you. Text transcriptions of these recorded calls can also be produced.

Recordings remain stored indefinitely. To delete a recording, use the appropriate API call from the [Compatibility API](/docs/compatibility-api/rest/recordings/delete-recording).

Any instructions placed after the `<Record>` verb will not be executed. To ensure additional instructions are processed, use the `action` attribute to specify a URL that SignalWire will request once the recording is complete.

## Verb attributes

<ParamField path="action" type="string">
  The `action` attribute takes in an absolute or relative URL. SignalWire will make a `GET` or `POST` request to this URL when recording is completed. The current document's URL will be requested if no `action` is provided which can lead to unwanted looping behavior if you're not careful. See [below](#record_action) for specified request parameters.
</ParamField>

<ParamField path="finishOnKey" type="string">
  The set of digits, (0-9, \*, #), that can end a recording.
</ParamField>

<ParamField path="maxLength" type="integer">
  The maximum length, in seconds, of the recording. A value of zero means 3600 seconds.
</ParamField>

<ParamField path="method" type="string" default="POST">
  The `method` attribute specifies whether the request to action is a `GET` or a `POST`. Valid values are `GET` or `POST`.
</ParamField>

<ParamField path="playBeep" type="boolean" default="true">
  Whether or not a sound is played before the start of a recording.
</ParamField>

<ParamField path="recordingStatusCallback" type="string">
  The `recordingStatusCallback` attribute takes in an absolute or relative URL. SignalWire will make a `GET` or `POST` request to this URL when recording is accessible. See [below](#record_recordingStatusCallback) for specified request parameters.
</ParamField>

<ParamField path="recordingStatusCallbackEvent" type="string" default="completed">
  The different recording statuses. Possible values are `completed`, `in-progress`, and `absent`. To specify multiple events, separate with a space.
</ParamField>

<ParamField path="recordingStatusCallbackMethod" type="string" default="POST">
  The type of HTTP request to use when requesting a `recordingStatusCallback`.
</ParamField>

<ParamField path="storageUrl" type="string">
  The `storageUrl` attribute accepts an absolute URL as the destination to send a recording to, if you prefer to host your own recordings and bypass SignalWire storage.
</ParamField>

<ParamField path="storageUrlMethod" type="string" default="POST">
  Specifies which HTTP verb to use when sending the recording to the `storageUrl`. Available values are: **POST** and **PUT**.
</ParamField>

<ParamField path="timeout" type="integer">
  The `timeout` attribute specifies the number of seconds of silence that ends a recording.
</ParamField>

<ParamField path="transcribe" type="boolean" default="false">
  The `transcribe` attribute identifies whether to produce a text transcription of the recording. There is an additional charge for this service, so is turned off by default.
</ParamField>

<ParamField path="transcribeCallback" type="string">
  The ability to define a URL to which SignalWire will make a `POST` request to once the transcription is complete. See [below](#record_transcribeCallback) for specified request parameters.
</ParamField>

<ParamField path="trim" type="string" default="trim-silence">
  Whether or not silence in the beginning and end of recordings are removed. Allowed values are `trim-silence` and `do-not-trim`.
</ParamField>

<Info title="the sound of silence">
  If no audio data is received, including when a caller is silent and `trim-silence` is enabled, SignalWire will not save a recording.
  If you wish to save silence, be sure to set `trim="do-not-trim"`.

  Note also that SignalWire will trim leading and trailing silence from your audio files, causing the duration of calls to be less than the time spent recording.
</Info>

<Warning title="avoid looping">
  When recording finishes, including when no audio data is received, `<Record>` will always request its `action` URL and process the cXML instructions that are returned.
  If no `action` URL is set, SignalWire will re-request the current cXML document's URL by default.
  **This can lead to unwanted looping behavior**, so make sure to end the call using `action` as seen in the
  [Recording a Voicemail example](#recording-a-voicemail).
</Warning>

#### Request parameters for `action` URL  \[#record\_action]

The `action` request contains the [Standard Request Parameters](/docs/compatibility-api/cxml/reference/voice#request-parameters) as well as:

<ParamField path="Digits" type="string">
  The buttons pressed to end a recording.
</ParamField>

<ParamField path="RecordingDuration" type="integer">
  The duration, in seconds, of the audio recording.
</ParamField>

<ParamField path="RecordingUrl" type="string">
  The URL of the recorded audio file.
</ParamField>

#### Request parameters for `recordingStatusCallback`  \[#record\_recordingStatusCallback]

<WebhookPayloadSnippet webhook="recordingStatusCallback" />

#### Request parameters for `transcribeCallback`  \[#record\_transcribeCallback]

<WebhookPayloadSnippet webhook="transcriptionStatusCallback" />

## Nesting

No other verbs can be nested within `<Record>` and you cannot nest `<Record>` within any other verbs.

## Examples

### Recording a voicemail

<CodeBlocks>
  ```xml
  <?xml version="1.0" encoding="UTF-8"?>
  <Response>
      <Say>
          Please leave a message at the beep.
          Press the pound key when finished.
      </Say>
      <Record
          action="http://your-application.com/handleRecording.php"
          method="GET"
          maxLength="15"
          finishOnKey="#"
          />
  </Response>
  ```

  ```javascript title="Node.js"
  const { RestClient } = require("@signalwire/compatibility-api");
  const response = new RestClient.LaML.VoiceResponse();

  response.say("Please leave a message at the beep. Press the pound key when finished.");
  response.record({
    action: "http://your-application.com/handleRecording.php",
    method: "GET",
    maxLength: 15,
    finishOnKey: "#",
  });
  console.log(response.toString());
  ```

  ```csharp
  using Twilio.TwiML;
  using Twilio.Http;
  using System;


  class Example
  {
      static void Main()
      {
          var response = new VoiceResponse();
          response.Say("Please leave a message at the beep. Press the pound key when finished.");
          response.Record(action: new Uri("http://your-application.com/handleRecording.php"),
              method: HttpMethod.Get, maxLength: 15, finishOnKey: "#");

          Console.WriteLine(response.ToString());;
      }
  }
  ```

  ```python
  from signalwire.voice_response import VoiceResponse, Record, Say

  response = VoiceResponse()
  response.say('Please leave a message at the beep. Press the pound key when finished.')
  response.record(action='http://your-application.com/handleRecording.php', method='GET', max_length=15, finish_on_key='#')

  print(response)
  ```

  ```ruby
  require 'signalwire/sdk'

  response = Signalwire::Sdk::VoiceResponse.new do |response|
    response.say(message: 'Please leave a message at the beep. Press the pound key when finished.')
    response.record(action: 'http://your-application.com/handleRecording.php', method: 'GET', max_length: 15, finish_on_key: '#')
  end

  puts response.to_s
  ```
</CodeBlocks>

This prompt will play before the 'beep', asking the caller to leave a message. The caller can only leave a message that is 15s long.

### Transcribing a recording

<CodeBlocks>
  ```xml
  <?xml version="1.0" encoding="UTF-8"?>
  <Response>
      <Record
          transcribe="true"
          transcribeCallback="http://your-application.com/handle_transcribe.php" />
  </Response>
  ```

  ```javascript title="Node.js"
  const { RestClient } = require("@signalwire/compatibility-api");
  const response = new RestClient.LaML.VoiceResponse();

  response.record({
    transcribe: true,
    transcribeCallback: "http://your-application.com/handle_transcribe.php",
  });
  console.log(response.toString());
  ```

  ```csharp
  using Twilio.TwiML;
  using System;


  class Example
  {
      static void Main()
      {
          var response = new VoiceResponse();
          response.Record(transcribe: true,
              transcribeCallback: new Uri("http://your-application.com/handle_transcribe.php"));

          Console.WriteLine(response.ToString());;
      }
  }
  ```

  ```python
  from signalwire.voice_response import VoiceResponse, Record

  response = VoiceResponse()
  response.record(transcribe=True, transcribe_callback='http://your-application.com/handle_transcribe.php')

  print(response)
  ```

  ```ruby
  require 'signalwire/sdk'

  response = Signalwire::Sdk::VoiceResponse.new do |response|
    response.record(transcribe: true, transcribe_callback: 'http://your-application.com/handle_transcribe.php')
  end

  puts response.to_s
  ```
</CodeBlocks>

SignalWire will record the caller and transcribe the recording once it is complete.
Then, SignalWire will make a `POST` request to the `transcribeCallback` URL with the transcription as a parameter.