stts 1.3.4 copy "stts: ^1.3.4" to clipboard
stts: ^1.3.4 copied to clipboard

Published 41 hours ago • verified publishercow-level.ovh
SDKFlutter
PlatformAndroidiOSmacOSwebWindows
→

Metadata

Speech-to-Text and Text-to-Speech plugin. Offline first.

More...

1.3.4 #

  • chore(ios): Require iOS 13, like the oldest supported Flutter version.
  • fix(macos): Dispose STT and TTS when the app terminates.
  • refactor(darwin): Share iOS and macOS sources, macOS gets all iOS fixes below.
  • fix(ios): Normalize language tags (e.g. en_US, en-us).
  • fix(ios): Return an error instead of crashing on invalid TTS arguments.
  • fix(ios): Ignore TTS finish events of flushed or stopped utterances.
  • fix(ios): Send last STT partial result as final on stop.
  • fix(ios): Don't read STT request from the audio thread.
  • fix(ios): Default STT language to a supported one.
  • fix(ios): Ignore STT start while listening.
  • fix(ios): Release STT resources when start fails.
  • fix(ios): Report STT error instead of crashing when audio input is not available.
  • fix(ios): Keep TTS voice and language in sync.
  • fix(ios): Restore the audio session on STT stop when the plugin manages it.
  • fix(android): Don't mute ringer on devices without vibrator.
  • fix(android): Ignore TTS resume when not paused.
  • fix(android): Stop STT and restore ringer mode when recognition becomes unavailable.
  • fix(android): Track TTS resume position per utterance.
  • fix(android): Normalize language tags (e.g. en_US, en-us).
  • fix(android): Report download model end when on device recognition is unavailable.
  • fix(android): Retrieve STT languages only once during initialization.
  • fix: Queue all calls, including isSupported and platform specific ones.
  • fix(android): Guard TTS init against concurrent dispose.
  • fix(android): Run speech recognizer callbacks on main thread.
  • fix(android): Mutate TTS utterance data on main thread only.
  • fix(android): Restore ringer mode on quick stop/start cycle.
  • fix(Android): Disable offline mode when no language is retrieved.
  • fix(windows): Initialize TTS option values.
  • fix(windows): Escape TTS text as XML.
  • fix(windows): Release audio input token on STT start.
  • fix(windows): Select first matching voice on language change.
  • fix(windows): Keep STT recognition context alive until dispose.
  • fix(windows): Return empty list when TTS voices cannot be enumerated.
  • fix(windows): Encode SAPI strings as UTF-8.
  • fix(windows): Release SAPI interfaces on error paths.
  • fix(windows): Allow STT restart right after stop.
  • fix(windows): Identify TTS voices by token id.
  • fix(windows): Handle all queued STT events, drop the ones of a stopped session.

1.3.3 #

  • fix(ios): allowBluetooth deprecation.
  • fix(darwin): Properly propagate SttError errors.
  • fix(darwin): Dissociate state check vs. listener check.
  • fix(darwin): TO now provides final result.
  • fix(darwin): Remove now useless tiny timer on iOS when stopping with fix applied in 1.3.2.

1.3.2 #

  • fix(darwin): prevent tap reinstall crash on rapid start/stop.

1.3.1 #

  • fix(Android): replace force-unwrap of TTS start arguments with callOrError.
  • fix(android): Prevent missed TtsState.Stop.
  • fix(Android): Potential NPE.
  • fix(Windows): Pointer release issues.
  • fix(darwin): transcription.segments[0] without empty check crashes on empty segments.
  • fix(darwin): Don't Apple's main-thread requirement for AVSpeechSynthesizer.

1.3.0 #

  • chore: Updates minimum supported SDK version to Flutter 3.44/Dart 3.12.
  • Android:
    • chore: Move to AGP 9.x.
    • chore: Move to Kotlin Gradle DSL.
  • iOS:
    • chore: Completes Swift Package Manager integration.
  • macOS:
    • chore: Completes Swift Package Manager integration.

1.2.6 #

  • fix(Android): Check if on device recognition is available before creating.

1.2.5 #

  • fix(Android/STT): Improve compatibility with older APIs/devices and offline mode.
  • chore(Android): compileSdk 36 / kotlin 2.2.

1.2.4 #

  • fix(Android/STT): Offline/online behaviour consistency (and truncated detection).

1.2.3 #

  • fix(iOS/macOS/TTS): Event firing regression.

1.2.2 #

  • fix(iOS/macOS/TTS): Event firing regression.

1.2.1 #

  • fix(iOS/macOS/TTS): Event firing regression.

1.2.0 #

  • feat(STT): Allow online usage.
  • fix(iOS/macOS): compilation issue under some circumstances.
  • fix(iOS/macOS/TTS): Event firing when stressing the plugin.
  • fix(iOS/macOS/TTS): Stay away from UI thread.

1.1.2 #

  • fix(iOS/macOS/TTS): Event firing when stressing the plugin.

1.1.1 #

  • fix(iOS/macOS/TTS): Event firing when stressing the plugin.

1.1.0 #

  • feat(TTS): Add queue flush option.
  • feat(TTS): Add pre/post silence delay option.
  • fix(Android/STT): Increase delay by 100ms for muteSystemSounds. Ends of sounds were sometimes hearable.

1.0.2 #

  • fix(iOS/macOS/STT): Properly initialize default language.

1.0.1 #

  • fix(Android/TTS): Double implicit cast error.
  • fix(iOS/TTS): Crash when setting voice.
  • fix(macOS/TTS): Crash when setting voice.

1.0.0 #

  • feat(iOS/STT): Add audio session management.
  • fix(Windows): Throw PlatformException on error.
  • fix(Android/STT): Code cleanup with some improvement.
  • fix(iOS/macOS): Small code fixes/improvements.

0.9.1 #

  • feat(Android/STT): Add muteSystemSounds to mute default system beep sounds.
  • feat(STT): Add recognition options.

0.9.0 #

  • Supports all platforms but linux.
  • Beta stage.

← Metadata

39
likes
160
points
2.01k
downloads

Documentation

API reference

Publisher

verified publishercow-level.ovh

Weekly Downloads

Metadata

Speech-to-Text and Text-to-Speech plugin. Offline first.

Repository (GitHub)
View/report issues

Topics

#speech-to-text #speech-recognition #text-to-speech #tts #voice-synthesizer

License

BSD-3-Clause (license)

Dependencies

flutter, stts_platform_interface, stts_web

More

Packages that depend on stts

Packages that implement stts

Back