AndroidWorld Benchmark

Droidrun achieves state of the art performance on AndroidWorld with 78.4% success rate across 116 diverse tasks. Despite setup challenges and evaluation limitations, our open source approach using GPT-5 and direct Accessibility API access outperforms all competing agents.

AndroidWorld Benchmark Results

Success rates of leading AI agents on the 116-task AndroidWorld benchmark (18.09.2025)

Droidrun
Other Agents
78.4% Success Rate
AndroidWorld benchmark results
91/116 (78.4%)
91 Success
25 Failed

Audio Recorder Record Audio

Task: AudioRecorderRecordAudio

Success

Audio Recorder Record Audio With File Name

Task: AudioRecorderRecordAudioWithFileName

Success

Browser Maze

Task: BrowserMaze

Success

Camera Take Photo

Task: CameraTakePhoto

Success

Camera Take Video

Task: CameraTakeVideo

Success

Clock Stop Watch Paused Verify

Task: ClockStopWatchPausedVerify

Success

Clock Stop Watch Running

Task: ClockStopWatchRunning

Success

Clock Timer Entry

Task: ClockTimerEntry

Success

Contacts Add Contact

Task: ContactsAddContact

Success

Contacts New Contact Draft

Task: ContactsNewContactDraft

Success

Expense Add Multiple

Task: ExpenseAddMultiple

Success

Expense Add Multiple From Gallery

Task: ExpenseAddMultipleFromGallery

Success

Expense Add Single

Task: ExpenseAddSingle

Success

Expense Delete Duplicates

Task: ExpenseDeleteDuplicates

Success

Expense Delete Duplicates2

Task: ExpenseDeleteDuplicates2

Success

Expense Delete Multiple

Task: ExpenseDeleteMultiple

Success

Expense Delete Multiple2

Task: ExpenseDeleteMultiple2

Success

Expense Delete Single

Task: ExpenseDeleteSingle

Success

Files Delete File

Task: FilesDeleteFile

Success

Files Move File

Task: FilesMoveFile

Success

Markor Create Folder

Task: MarkorCreateFolder

Success

Markor Create Note

Task: MarkorCreateNote

Success

Markor Create Note And Sms

Task: MarkorCreateNoteAndSms

Success

Markor Create Note From Clipboard

Task: MarkorCreateNoteFromClipboard

Success

Markor Delete All Notes

Task: MarkorDeleteAllNotes

Success

Markor Delete Newest Note

Task: MarkorDeleteNewestNote

Success

Markor Delete Note

Task: MarkorDeleteNote

Success

Markor Transcribe Receipt

Task: MarkorTranscribeReceipt

Success

Notes Is Todo

Task: NotesIsTodo

Success

Notes Meeting Attendee Count

Task: NotesMeetingAttendeeCount

Success

Notes Recipe Ingredient Count

Task: NotesRecipeIngredientCount

Success

Notes Todo Item Count

Task: NotesTodoItemCount

Success

Open App Task Eval

Task: OpenAppTaskEval

Success

Osm And Favorite

Task: OsmAndFavorite

Success

Recipe Add Multiple Recipes

Task: RecipeAddMultipleRecipes

Success

Recipe Add Single Recipe

Task: RecipeAddSingleRecipe

Success

Recipe Delete Duplicate Recipes

Task: RecipeDeleteDuplicateRecipes

Success

Recipe Delete Multiple Recipes

Task: RecipeDeleteMultipleRecipes

Success

Recipe Delete Multiple Recipes With Constraint

Task: RecipeDeleteMultipleRecipesWithConstraint

Success

Recipe Delete Multiple Recipes With Noise

Task: RecipeDeleteMultipleRecipesWithNoise

Success

Recipe Delete Single Recipe

Task: RecipeDeleteSingleRecipe

Success

Recipe Delete Single With Recipe With Noise

Task: RecipeDeleteSingleWithRecipeWithNoise

Success

Retro Create Playlist

Task: RetroCreatePlaylist

Success

Retro Playing Queue

Task: RetroPlayingQueue

Success

Retro Playlist Duration

Task: RetroPlaylistDuration

Success

Retro Save Playlist

Task: RetroSavePlaylist

Success

Save Copy Of Receipt Task Eval

Task: SaveCopyOfReceiptTaskEval

Success

Simple Calendar Add One Event

Task: SimpleCalendarAddOneEvent

Success

Simple Calendar Add One Event In Two Weeks

Task: SimpleCalendarAddOneEventInTwoWeeks

Success

Simple Calendar Add One Event Relative Day

Task: SimpleCalendarAddOneEventRelativeDay

Success

Simple Calendar Add One Event Tomorrow

Task: SimpleCalendarAddOneEventTomorrow

Success

Simple Calendar Add Repeating Event

Task: SimpleCalendarAddRepeatingEvent

Success

Simple Calendar Delete Events

Task: SimpleCalendarDeleteEvents

Success

Simple Calendar Delete Events On Relative Day

Task: SimpleCalendarDeleteEventsOnRelativeDay

Success

Simple Calendar Delete One Event

Task: SimpleCalendarDeleteOneEvent

Success

Simple Calendar Event On Date At Time

Task: SimpleCalendarEventOnDateAtTime

Success

Simple Calendar Events In Time Range

Task: SimpleCalendarEventsInTimeRange

Success

Simple Calendar Events On Date

Task: SimpleCalendarEventsOnDate

Success

Simple Calendar First Event After Start Time

Task: SimpleCalendarFirstEventAfterStartTime

Success

Simple Calendar Location Of Event

Task: SimpleCalendarLocationOfEvent

Success

Simple Calendar Next Event

Task: SimpleCalendarNextEvent

Success

Simple Calendar Next Meeting With Person

Task: SimpleCalendarNextMeetingWithPerson

Success

Simple Draw Pro Create Drawing

Task: SimpleDrawProCreateDrawing

Success

Simple Sms Reply

Task: SimpleSmsReply

Success

Simple Sms Reply Most Recent

Task: SimpleSmsReplyMostRecent

Success

Simple Sms Resend

Task: SimpleSmsResend

Success

Simple Sms Send

Task: SimpleSmsSend

Success

Simple Sms Send Clipboard Content

Task: SimpleSmsSendClipboardContent

Success

Simple Sms Send Received Address

Task: SimpleSmsSendReceivedAddress

Success

Sports Tracker Activities Count For Week

Task: SportsTrackerActivitiesCountForWeek

Success

Sports Tracker Activity Duration

Task: SportsTrackerActivityDuration

Success

Sports Tracker Total Duration For Category This Week

Task: SportsTrackerTotalDurationForCategoryThisWeek

Success

System Bluetooth Turn Off

Task: SystemBluetoothTurnOff

Success

System Bluetooth Turn Off Verify

Task: SystemBluetoothTurnOffVerify

Success

System Bluetooth Turn On

Task: SystemBluetoothTurnOn

Success

System Bluetooth Turn On Verify

Task: SystemBluetoothTurnOnVerify

Success

System Brightness Max

Task: SystemBrightnessMax

Success

System Brightness Max Verify

Task: SystemBrightnessMaxVerify

Success

System Brightness Min

Task: SystemBrightnessMin

Success

System Brightness Min Verify

Task: SystemBrightnessMinVerify

Success

System Copy To Clipboard

Task: SystemCopyToClipboard

Success

System Wifi Turn Off

Task: SystemWifiTurnOff

Success

System Wifi Turn Off Verify

Task: SystemWifiTurnOffVerify

Success

System Wifi Turn On

Task: SystemWifiTurnOn

Success

System Wifi Turn On Verify

Task: SystemWifiTurnOnVerify

Success

Tasks Due On Date

Task: TasksDueOnDate

Success

Tasks High Priority Tasks

Task: TasksHighPriorityTasks

Success

Tasks Incomplete Tasks On Date

Task: TasksIncompleteTasksOnDate

Success

Turn Off Wifi And Turn On Bluetooth

Task: TurnOffWifiAndTurnOnBluetooth

Success

Turn On Wifi And Open App

Task: TurnOnWifiAndOpenApp

Success

Vlc Create Playlist

Task: VlcCreatePlaylist

Success

Browser Multiply

Task: BrowserMultiply

Failed

Expense Add Multiple From Markor

Task: ExpenseAddMultipleFromMarkor

Failed

Markor Add Note Header

Task: MarkorAddNoteHeader

Failed

Markor Change Note Content

Task: MarkorChangeNoteContent

Failed

Markor Edit Note

Task: MarkorEditNote

Failed

Markor Merge Notes

Task: MarkorMergeNotes

Failed

Markor Move Note

Task: MarkorMoveNote

Failed

Markor Transcribe Video

Task: MarkorTranscribeVideo

Failed

Osm And Marker

Task: OsmAndMarker

Failed

Osm And Track

Task: OsmAndTrack

Failed

Recipe Add Multiple Recipes From Image

Task: RecipeAddMultipleRecipesFromImage

Failed

Recipe Add Multiple Recipes From Markor

Task: RecipeAddMultipleRecipesFromMarkor

Failed

Recipe Add Multiple Recipes From Markor2

Task: RecipeAddMultipleRecipesFromMarkor2

Failed

Recipe Delete Duplicate Recipes2

Task: RecipeDeleteDuplicateRecipes2

Failed

Recipe Delete Duplicate Recipes3

Task: RecipeDeleteDuplicateRecipes3

Failed

Simple Calendar Any Events On Date

Task: SimpleCalendarAnyEventsOnDate

Failed

Simple Calendar Events In Next Week

Task: SimpleCalendarEventsInNextWeek

Failed

Sports Tracker Activities On Date

Task: SportsTrackerActivitiesOnDate

Failed

Sports Tracker Longest Distance Activity

Task: SportsTrackerLongestDistanceActivity

Failed

Sports Tracker Total Distance For Category Over Interval

Task: SportsTrackerTotalDistanceForCategoryOverInterval

Failed

Tasks Completed Tasks For Date

Task: TasksCompletedTasksForDate

Failed

Tasks Due Next Week

Task: TasksDueNextWeek

Failed

Tasks High Priority Tasks Due On Date

Task: TasksHighPriorityTasksDueOnDate

Failed

Vlc Create Two Playlists

Task: VlcCreateTwoPlaylists

Failed