Speecx is a free speech synthesis (TTS) dubbing software. It supports English, Japanese, French, Russian, German, Hindi, Italian, Thai, Korean, Spanish, Catalan, Malay, Arabic, Danish, Dutch, Finnish, Icelandic, Norwegian, Polish, Portuguese, Romanian, Swedish, Turkish, Welsh, and some offline languages for free use. It also provides a variety of natural readers to choose from.
The speech service enables applications to read on-screen text aloud. For example, it can be used for:
- Google Play Books to “read aloud” your favorite books.
- Google Translate, to speak the translation aloud so you can hear the words pronounced.
- TalkBack and accessibility apps that provide spoken feedback on the device.
- Supports IPA by adding a mark
[=](=IPA) after the word, such aspresent[=ˈpreznt]orpresent[=prɪˈzent]. - Download and share any text to audio (MP3/AAC/FLAC).
- Powers the AI speechify engine for many other apps in the Play Store.
How to Use
To use the Speecx text-to-speech feature on your Android device:
- Go to Settings > Language & input > Text-to-speech output.
- Choose Speecx as your engine of choice.
You can add background music for mixing. Download the text-to-audio file to your device and share it.
View on Google Play StoreFree or paid:
- Free with in-app purchase
Accessibility and User Comments:
Upon launching the app for the first time, you will be asked to log in or skip and use the offline voices. If you choose to use offline voices, you will be presented with a list of available options. You can press the play button to hear a sample of each voice. Next to the play button is the download button, which is unlabeled; however, TalkBack's label detection should provide a meaningful label. After downloading a voice, the button changes to a "Select" button, allowing you to set that voice as the default.
Tapping the name of the selected voice in the Home tab will take you to the full list of voices, where you can select the voices you want. Downloaded voices appear under a "Downloaded" item, where you can delete them using the Delete button.
The app also supports exporting audio files synthesized within the app and mixing speech with music, but I haven’t tested these features, as my focus is solely on the system TTS engine functionality. This engine can be set as the preferred system engine, which the screen reader could then utilize.
It’s important to make sure you’ve selected the right voice in the app’s settings (accessible from the Home tab) before setting the engine as your system default to avoid losing speech output. The settings also include speed and volume controls, but tapping on these doesn’t seem to have any effect.
The quality of the offline voices can’t be compared to options like Acapela or Vocalizer, but they are fairly responsive. The range of offline-supported voices and languages is limited. I haven’t subscribed to any of the available plans (weekly or monthly), as I found them quite expensive. It appears that all subscription voices are online-only, which could be a dealbreaker for many screen reader users. The engine is expected to fall back to a downloaded offline voice if the connection is lost.
With the free plan, you can still test premium voices, but only with a limited daily quota. Also, the free version includes many ads. You must watch a full-screen ad to download a voice, and opening the app typically displays another. Ads are also embedded throughout the app's interface.
Almost every button in the app is unlabeled, though TalkBack’s label detection usually does the heavy lifting, and buttons can be manually labeled. The UI is clunky and overwhelming—it could benefit from simplification—but it remains navigable.
Additionally, it’s worth mentioning that even if auto-rotation is locked to portrait, the app doesn’t obey this setting; instead, it keeps switching between landscape and portrait modes based on how you move the phone.
I don’t consider this TTS engine something I’d use as my default, but it’s good to have as a backup if you’re a TTS junkie like me who likes to change things up from time to time. I tested it with article reading and it worked well, though there were some pronunciation oddities here and there.
You can listen to my audio demo of the engine and its UI here.
Last Tested App version and Android version:
App version: 1.24.0821, Android 16
Comments