Added in API level 8

SpeechRecognizer


open class SpeechRecognizer
kotlin.Any
   ↳ android.speech.SpeechRecognizer

This class provides access to the speech recognition service. This service allows access to the speech recognizer. Do not instantiate this class directly, instead, call SpeechRecognizer#createSpeechRecognizer(Context), or SpeechRecognizer#createOnDeviceSpeechRecognizer(Context). This class's methods must be invoked only from the main application thread.

Important: the caller MUST invoke destroy() on a SpeechRecognizer object when it is no longer needed.

The implementation of this API is likely to stream audio to remote servers to perform speech recognition. As such this API is not intended to be used for continuous recognition, which would consume a significant amount of battery and bandwidth.

Please note that the application must have android.Manifest.permission#RECORD_AUDIO permission to use this class.

Summary

Constants
static String

Key used to retrieve a float array from the Bundle passed to the RecognitionListener#onResults(Bundle) and RecognitionListener#onPartialResults(Bundle) methods.

static String

Key used to retrieve a String representation of the IETF language tag (as defined by BCP 47, e.g., "en-US", "de-DE") of the detected language of the most recent audio chunk.

static Int

Audio recording error.

static Int

The service does not allow to check for support.

static Int

The service does not support listening to model downloads events.

static Int

Other client side errors.

static Int

Insufficient permissions

static Int

Requested language is not available to be used with the current recognizer.

static Int

Requested language is supported, but not available currently (e.g. not downloaded yet).

static Int

Other network related errors.

static Int

Network operation timed out.

static Int

No recognition result matched.

static Int

RecognitionService busy.

static Int

Server sends error status.

static Int

Server has been disconnected, e.

static Int

No speech input

static Int

Too many requests from the same client.

static String

Key used to retrieve the level of confidence of the detected language of the most recent audio chunk, represented by an int value prefixed by LANGUAGE_DETECTION_CONFIDENCE_LEVEL_.

static Int

static Int

static Int

static Int

static String

Key used to retrieve the result of the language switch of the most recent audio chunk, represented by an int value prefixed by LANGUAGE_SWITCH_.

static Int

Switch attempted and failed.

static Int

Switch not attempted.

static Int

Switch skipped because the language model is missing or the language is not allowlisted for auto switch.

static Int

Switch attempted and succeeded.

static String

Key used to receive an ArrayList<RecognitionPart> object from the Bundle passed to the RecognitionListener#onResults(Bundle) and RecognitionListener#onSegmentResults(Bundle) methods.

static String

Key used to retrieve an ArrayList<AlternativeSpans> from the Bundle passed to the RecognitionListener#onResults(Bundle) and RecognitionListener#onPartialResults(Bundle) methods.

static String

Key used to retrieve an ArrayList<String> from the Bundle passed to the RecognitionListener#onResults(Bundle) and RecognitionListener#onPartialResults(Bundle) methods.

static String

Key used to retrieve an ArrayList<String> containing representations of the IETF language tags (as defined by BCP 47, e.g., "en-US", "en-UK") denoting the alternative locales for the same language retrieved by the key

Public methods
open Unit

Cancels the speech recognition.

open Unit
checkRecognitionSupport(recognizerIntent: Intent, executor: Executor, supportListener: RecognitionSupportCallback)

Checks whether recognizerIntent is supported by SpeechRecognizer#startListening(Intent).

open static SpeechRecognizer

Factory method to create a new SpeechRecognizer.

open static SpeechRecognizer!

Factory method to create a new SpeechRecognizer.

open static SpeechRecognizer!
createSpeechRecognizer(context: Context!, serviceComponent: ComponentName!)

Factory method to create a new SpeechRecognizer.

open Unit

Destroys the SpeechRecognizer object.

open static Boolean

Checks whether an on-device speech recognition service is available on the system.

open static Boolean

Checks whether a speech recognition service is available on the system.

open Unit

Sets the listener that will receive all the callbacks.

open Unit
startListening(recognizerIntent: Intent!)

Starts listening for speech.

open Unit

Stops listening for speech.

open Unit
triggerModelDownload(recognizerIntent: Intent)

Attempts to download the support for the given recognizerIntent.

open Unit
triggerModelDownload(recognizerIntent: Intent, executor: Executor, listener: ModelDownloadListener)

Attempts to download the support for the given recognizerIntent.

Constants

CONFIDENCE_SCORES

Added in API level 14
static val CONFIDENCE_SCORES: String

Key used to retrieve a float array from the Bundle passed to the RecognitionListener#onResults(Bundle) and RecognitionListener#onPartialResults(Bundle) methods. The array should be the same size as the ArrayList provided in RESULTS_RECOGNITION, and should contain values ranging from 0.0 to 1.0, or -1 to represent an unavailable confidence score.

Confidence values close to 1.0 indicate high confidence (the speech recognizer is confident that the recognition result is correct), while values close to 0.0 indicate low confidence.

This value is optional and might not be provided.

Value: "confidence_scores"

DETECTED_LANGUAGE

Added in API level 34
static val DETECTED_LANGUAGE: String

Key used to retrieve a String representation of the IETF language tag (as defined by BCP 47, e.g., "en-US", "de-DE") of the detected language of the most recent audio chunk.

This info is returned to the client in the Bundle passed to RecognitionListener#onLanguageDetection(Bundle) only if RecognizerIntent#EXTRA_ENABLE_LANGUAGE_DETECTION is set. Additionally, if RecognizerIntent#EXTRA_LANGUAGE_DETECTION_ALLOWED_LANGUAGES are listed, the detected language is constrained to be one from the list.

Value: "detected_language"

ERROR_AUDIO

Added in API level 8
static val ERROR_AUDIO: Int

Audio recording error.

Value: 3

ERROR_CANNOT_CHECK_SUPPORT

Added in API level 33
static val ERROR_CANNOT_CHECK_SUPPORT: Int

The service does not allow to check for support.

Value: 14

ERROR_CANNOT_LISTEN_TO_DOWNLOAD_EVENTS

Added in API level 34
static val ERROR_CANNOT_LISTEN_TO_DOWNLOAD_EVENTS: Int

The service does not support listening to model downloads events.

Value: 15

ERROR_CLIENT

Added in API level 8
static val ERROR_CLIENT: Int

Other client side errors.

Value: 5

ERROR_INSUFFICIENT_PERMISSIONS

Added in API level 8
static val ERROR_INSUFFICIENT_PERMISSIONS: Int

Insufficient permissions

Value: 9

ERROR_LANGUAGE_NOT_SUPPORTED

Added in API level 31
static val ERROR_LANGUAGE_NOT_SUPPORTED: Int

Requested language is not available to be used with the current recognizer.

Value: 12

ERROR_LANGUAGE_UNAVAILABLE

Added in API level 31
static val ERROR_LANGUAGE_UNAVAILABLE: Int

Requested language is supported, but not available currently (e.g. not downloaded yet).

Value: 13

ERROR_NETWORK

Added in API level 8
static val ERROR_NETWORK: Int

Other network related errors.

Value: 2

ERROR_NETWORK_TIMEOUT

Added in API level 8
static val ERROR_NETWORK_TIMEOUT: Int

Network operation timed out.

Value: 1

ERROR_NO_MATCH

Added in API level 8
static val ERROR_NO_MATCH: Int

No recognition result matched.

Value: 7

ERROR_RECOGNIZER_BUSY

Added in API level 8
static val ERROR_RECOGNIZER_BUSY: Int

RecognitionService busy.

Value: 8

ERROR_SERVER

Added in API level 8
static val ERROR_SERVER: Int

Server sends error status.

Value: 4

ERROR_SERVER_DISCONNECTED

Added in API level 31
static val ERROR_SERVER_DISCONNECTED: Int

Server has been disconnected, e.g. because the app has crashed.

Value: 11

ERROR_SPEECH_TIMEOUT

Added in API level 8
static val ERROR_SPEECH_TIMEOUT: Int

No speech input

Value: 6

ERROR_TOO_MANY_REQUESTS

Added in API level 31
static val ERROR_TOO_MANY_REQUESTS: Int

Too many requests from the same client.

Value: 10

LANGUAGE_DETECTION_CONFIDENCE_LEVEL

Added in API level 34
static val LANGUAGE_DETECTION_CONFIDENCE_LEVEL: String

Key used to retrieve the level of confidence of the detected language of the most recent audio chunk, represented by an int value prefixed by LANGUAGE_DETECTION_CONFIDENCE_LEVEL_.

This info is returned to the client in the Bundle passed to RecognitionListener#onLanguageDetection(Bundle) only if RecognizerIntent#EXTRA_ENABLE_LANGUAGE_DETECTION is set.

Value: "language_detection_confidence_level"

LANGUAGE_DETECTION_CONFIDENCE_LEVEL_CONFIDENT

Added in API level 34
static val LANGUAGE_DETECTION_CONFIDENCE_LEVEL_CONFIDENT: Int
Value: 2

LANGUAGE_DETECTION_CONFIDENCE_LEVEL_HIGHLY_CONFIDENT

Added in API level 34
static val LANGUAGE_DETECTION_CONFIDENCE_LEVEL_HIGHLY_CONFIDENT: Int
Value: 3

LANGUAGE_DETECTION_CONFIDENCE_LEVEL_NOT_CONFIDENT

Added in API level 34
static val LANGUAGE_DETECTION_CONFIDENCE_LEVEL_NOT_CONFIDENT: Int
Value: 1

LANGUAGE_DETECTION_CONFIDENCE_LEVEL_UNKNOWN

Added in API level 34
static val LANGUAGE_DETECTION_CONFIDENCE_LEVEL_UNKNOWN: Int
Value: 0

LANGUAGE_SWITCH_RESULT

Added in API level 34
static val LANGUAGE_SWITCH_RESULT: String

Key used to retrieve the result of the language switch of the most recent audio chunk, represented by an int value prefixed by LANGUAGE_SWITCH_.

This info is returned to the client in the Bundle passed to the RecognitionListener#onLanguageDetection(Bundle) only if RecognizerIntent#EXTRA_ENABLE_LANGUAGE_SWITCH is set.

Value: "language_switch_result"

LANGUAGE_SWITCH_RESULT_FAILED

Added in API level 34
static val LANGUAGE_SWITCH_RESULT_FAILED: Int

Switch attempted and failed.

Value: 2

LANGUAGE_SWITCH_RESULT_NOT_ATTEMPTED

Added in API level 34
static val LANGUAGE_SWITCH_RESULT_NOT_ATTEMPTED: Int

Switch not attempted.

Value: 0

LANGUAGE_SWITCH_RESULT_SKIPPED_NO_MODEL

Added in API level 34
static val LANGUAGE_SWITCH_RESULT_SKIPPED_NO_MODEL: Int

Switch skipped because the language model is missing or the language is not allowlisted for auto switch.

Value: 3

LANGUAGE_SWITCH_RESULT_SUCCEEDED

Added in API level 34
static val LANGUAGE_SWITCH_RESULT_SUCCEEDED: Int

Switch attempted and succeeded.

Value: 1

RECOGNITION_PARTS

Added in API level 34
static val RECOGNITION_PARTS: String

Key used to receive an ArrayList<RecognitionPart> object from the Bundle passed to the RecognitionListener#onResults(Bundle) and RecognitionListener#onSegmentResults(Bundle) methods.

A single SpeechRecognizer result is represented as a String. Each word of the resulting String, as well as any potential adjacent punctuation, is represented by a RecognitionPart item from the ArrayList retrieved by this key.

Value: "recognition_parts"

RESULTS_ALTERNATIVES

Added in API level 34
static val RESULTS_ALTERNATIVES: String

Key used to retrieve an ArrayList<AlternativeSpans> from the Bundle passed to the RecognitionListener#onResults(Bundle) and RecognitionListener#onPartialResults(Bundle) methods. The list should be the same size as the ArrayList provided in RESULTS_RECOGNITION.

A single SpeechRecognizer result is represented as a String. For a specific span (substring) of the originally recognized result string the recognizer provides a list of alternative hypotheses in the form of an AlternativeSpan object. Alternatives for different spans of a result string are listed in an AlternativeSpans object. Each item from the ArrayList retrieved by this key corresponds to a single result string provided in RESULTS_RECOGNITION.

This value is optional and might not be provided.

Value: "results_alternatives"

RESULTS_RECOGNITION

Added in API level 8
static val RESULTS_RECOGNITION: String

Key used to retrieve an ArrayList<String> from the Bundle passed to the RecognitionListener#onResults(Bundle) and RecognitionListener#onPartialResults(Bundle) methods. These strings are the possible recognition results, where the first element is the most likely candidate.

Value: "results_recognition"

TOP_LOCALE_ALTERNATIVES

Added in API level 34
static val TOP_LOCALE_ALTERNATIVES: String

Key used to retrieve an ArrayList<String> containing representations of the IETF language tags (as defined by BCP 47, e.g., "en-US", "en-UK") denoting the alternative locales for the same language retrieved by the key DETECTED_LANGUAGE. This info is returned to the client in the Bundle passed to RecognitionListener#onLanguageDetection(Bundle) only if RecognizerIntent#EXTRA_ENABLE_LANGUAGE_DETECTION is set.

Value: "top_locale_alternatives"

Public methods

cancel

Added in API level 8
open fun cancel(): Unit

Cancels the speech recognition. Please note that setRecognitionListener(android.speech.RecognitionListener) should be called beforehand, otherwise no notifications will be received.
This method must be called from the main thread of your app.

checkRecognitionSupport

Added in API level 33
open fun checkRecognitionSupport(
    recognizerIntent: Intent,
    executor: Executor,
    supportListener: RecognitionSupportCallback
): Unit

Checks whether recognizerIntent is supported by SpeechRecognizer#startListening(Intent).

Parameters
recognizerIntent Intent: contains parameters for the recognition to be performed. The intent may also contain optional extras. See RecognizerIntent for the list of supported extras, any unlisted extra might be ignored. This value cannot be null.
supportListener RecognitionSupportCallback: the listener on which to receive the support query results. This value cannot be null.
executor Executor: This value cannot be null. Callback and listener events are dispatched through this Executor, providing an easy way to control which thread is used. To dispatch events through the main thread of your application, you can use Context.getMainExecutor(). Otherwise, provide an Executor that dispatches to an appropriate thread.

createOnDeviceSpeechRecognizer

Added in API level 31
open static fun createOnDeviceSpeechRecognizer(context: Context): SpeechRecognizer

Factory method to create a new SpeechRecognizer.

Please note that setRecognitionListener(android.speech.RecognitionListener) should be called before dispatching any command to the created SpeechRecognizer, otherwise no notifications will be received.

Important: the caller MUST invoke destroy() on a SpeechRecognizer object when it is no longer needed.
This method must be called from the main thread of your app.

Parameters
context Context: in which to create SpeechRecognizer This value cannot be null.
Return
SpeechRecognizer a new on-device SpeechRecognizer. This value cannot be null.
Exceptions
java.lang.UnsupportedOperationException iff isOnDeviceRecognitionAvailable(android.content.Context) is false

createSpeechRecognizer

Added in API level 8
open static fun createSpeechRecognizer(context: Context!): SpeechRecognizer!

Factory method to create a new SpeechRecognizer. Please note that setRecognitionListener(android.speech.RecognitionListener) should be called before dispatching any command to the created SpeechRecognizer, otherwise no notifications will be received.

Important: the caller MUST invoke destroy() on a SpeechRecognizer object when it is no longer needed.

For apps targeting Android 11 (API level 30) interaction with a speech recognition service requires element to be added to the manifest file:

<code>&lt;queries&gt;
    &lt;intent&gt;
      &lt;action
         android:name="android.speech.RecognitionService" /&gt;
    &lt;/intent&gt;
  &lt;/queries&gt;
  </code>

This method must be called from the main thread of your app.
Parameters
context Context!: in which to create SpeechRecognizer
Return
SpeechRecognizer! a new SpeechRecognizer

createSpeechRecognizer

Added in API level 8
open static fun createSpeechRecognizer(
    context: Context!,
    serviceComponent: ComponentName!
): SpeechRecognizer!

Factory method to create a new SpeechRecognizer. Please note that setRecognitionListener(android.speech.RecognitionListener) should be called before dispatching any command to the created SpeechRecognizer, otherwise no notifications will be received. Use this version of the method to specify a specific service to direct this SpeechRecognizer to.

Important: the caller MUST invoke destroy() on a SpeechRecognizer object when it is no longer needed.

Important: before calling this method, please check via android.content.pm.PackageManager#queryIntentServices(Intent, int) that serviceComponent actually exists and provides RecognitionService#SERVICE_INTERFACE. Normally you would not use this; call createSpeechRecognizer(android.content.Context) to use the system default recognition service instead or createOnDeviceSpeechRecognizer(android.content.Context) to use on-device recognition.

For apps targeting Android 11 (API level 30) interaction with a speech recognition service requires element to be added to the manifest file:

<code>&lt;queries&gt;
    &lt;intent&gt;
      &lt;action
         android:name="android.speech.RecognitionService" /&gt;
    &lt;/intent&gt;
  &lt;/queries&gt;
  </code>

This method must be called from the main thread of your app.
Parameters
context Context!: in which to create SpeechRecognizer
serviceComponent ComponentName!: the ComponentName of a specific service to direct this SpeechRecognizer to
Return
SpeechRecognizer! a new SpeechRecognizer

destroy

Added in API level 8
open fun destroy(): Unit

Destroys the SpeechRecognizer object.

isOnDeviceRecognitionAvailable

Added in API level 31
open static fun isOnDeviceRecognitionAvailable(context: Context): Boolean

Checks whether an on-device speech recognition service is available on the system. If this method returns false, SpeechRecognizer#createOnDeviceSpeechRecognizer(Context) will fail.

Parameters
context Context: with which on-device SpeechRecognizer will be created This value cannot be null.
Return
Boolean true if on-device recognition is available, false otherwise

isRecognitionAvailable

Added in API level 8
open static fun isRecognitionAvailable(context: Context): Boolean

Checks whether a speech recognition service is available on the system. If this method returns false, SpeechRecognizer#createSpeechRecognizer(Context) will fail.

Parameters
context Context: with which SpeechRecognizer will be created This value cannot be null.
Return
Boolean true if recognition is available, false otherwise

setRecognitionListener

Added in API level 8
open fun setRecognitionListener(listener: RecognitionListener!): Unit

Sets the listener that will receive all the callbacks. The previous unfinished commands will be executed with the old listener, while any following command will be executed with the new listener.
This method must be called from the main thread of your app.

Parameters
listener RecognitionListener!: listener that will receive all the callbacks from the created SpeechRecognizer, this must not be null.

startListening

Added in API level 8
open fun startListening(recognizerIntent: Intent!): Unit

Starts listening for speech. Please note that setRecognitionListener(android.speech.RecognitionListener) should be called beforehand, otherwise no notifications will be received.
This method must be called from the main thread of your app.

Parameters
recognizerIntent Intent!: contains parameters for the recognition to be performed. The intent may also contain optional extras, see RecognizerIntent. If these values are not set explicitly, default values will be used by the recognizer.

stopListening

Added in API level 8
open fun stopListening(): Unit

Stops listening for speech. Speech captured so far will be recognized as if the user had stopped speaking at this point.

Note that in the default case, this does not need to be called, as the speech endpointer will automatically stop the recognizer listening when it determines speech has completed. However, you can manipulate endpointer parameters directly using the intent extras defined in RecognizerIntent, in which case you may sometimes want to manually call this method to stop listening sooner.

Upon invocation clients must wait until RecognitionListener#onResults or RecognitionListener#onError are invoked before calling SpeechRecognizer#startListening again. Otherwise such an attempt would be rejected by recognition service.

Please note that setRecognitionListener(android.speech.RecognitionListener) should be called beforehand, otherwise no notifications will be received.
This method must be called from the main thread of your app.

triggerModelDownload

Added in API level 33
open fun triggerModelDownload(recognizerIntent: Intent): Unit

Attempts to download the support for the given recognizerIntent. This might trigger user interaction to approve the download. Callers can verify the status of the request via checkRecognitionSupport(android.content.Intent,java.util.concurrent.Executor,android.speech.RecognitionSupportCallback).

Parameters
recognizerIntent Intent: contains parameters for the recognition to be performed. The intent may also contain optional extras, see RecognizerIntent. This value cannot be null.

triggerModelDownload

Added in API level 34
open fun triggerModelDownload(
    recognizerIntent: Intent,
    executor: Executor,
    listener: ModelDownloadListener
): Unit

Attempts to download the support for the given recognizerIntent. This might trigger user interaction to approve the download. Callers can verify the status of the request via checkRecognitionSupport(android.content.Intent,java.util.concurrent.Executor,android.speech.RecognitionSupportCallback).

The updates about the model download request are received via the given ModelDownloadListener:

  • If the model is already available, ModelDownloadListener#onSuccess() will be called directly. The model can be safely used afterwards.
  • If the RecognitionService has started the download, ModelDownloadListener#onProgress(int) will be called an unspecified (zero or more) number of times until the download is complete. When the download finishes, ModelDownloadListener#onSuccess() will be called. The model can be safely used afterwards.
  • If the RecognitionService has only scheduled the download, but won't satisfy it immediately, ModelDownloadListener#onScheduled() will be called. There will be no further updates on this listener.
  • If the request fails at any time due to a network or scheduling error, ModelDownloadListener#onError(int) will be called.
  • Parameters
    recognizerIntent Intent: contains parameters for the recognition to be performed. The intent may also contain optional extras, see RecognizerIntent. This value cannot be null.
    executor Executor: for dispatching listener callbacks This value cannot be null. Callback and listener events are dispatched through this Executor, providing an easy way to control which thread is used. To dispatch events through the main thread of your application, you can use Context.getMainExecutor(). Otherwise, provide an Executor that dispatches to an appropriate thread.
    listener ModelDownloadListener: on which to receive updates about the model download request. This value cannot be null.