Table Of Contents
- Cloud Based Speech Services
- Native Android Text To Speech apps
- Speech Service by Google
- Cereplay Text To Speech Engine
- Vocalizer Text To Speech Engine
- Acapela Text To Speech Engine
- Samsung Text To Speech Engine
- Hear2Read Text To Speech Engine
- AhoTTS Text To Speech Engine
- Qfrency Text To Speech Engine
- Aharon Hebrew Text To Speech Engine
- SpeechLab 2.0 Text To Speech Engine
- SpeechTechTTS Text To Speech Engine
- SelvyTTS Text To Speech Engine
- RHVoice Text To Speech Engine
- Engines no longer available in Play Store
There are two types of Text To Speech engines you can use in Evie.
Cloud Based Services
These are not installed on your phone, and are only available over the internet.
They offer much higher quality than Android TTS engines.
They are paid services.
Native Android Text To Speech (TTS) engines.
These you can install from Play Store, and at least one already came with your phone.
They are free, and you can use them in Evie, for free.
Some of them require Internet, some of them work even when offline.
Cloud Based Speech Services
Evie offers 4 Cloud Based Speech Services:
- Amazon Polly
- Amazon Polly Neural
- Azure AI Speech
- Evie AI Speech
They are not installed on your phone, they are only available over the Internet. Evie allows you to use these services directly from the application, without any additional setup. There is no initial purchase, no installation or setup cost.
All Cloud Based Speech Services are paid. Prices are calculated by amount of text converted to speech so you only pay for what you use. To use them, you will need an Evie account and you will need to purchase voice credit.
Evie accepts Google Play payments in app to purchase credits.
Pricing model
Evie calculates how much voice you use in number of characters converted to speech.
1 million characters are roughly equivalent to 25 hours of speech, if you use the default voice settings.
That means a speech rate of 100% (to make Evie speak faster, you increase the speech rate).
The length of the generated audio will vary a lot. For example if you set the voices speed to 150%, the audio duration will probably be 17 hours.
Amazon Polly
Evie offers two types of Amazon Polly voices: Amazon Polly Standard and Amazon Polly Neural.
Amazon Polly Standard
Amazon Polly is a paid Text To Speech service that uses advanced deep learning technologies to synthesize speech that sounds like a human voice.
Amazon Polly Neural
Amazon Polly Neural is a paid Text To Speech service that uses advanced deep learning technologies to synthesize speech that sounds like a human voice. The Neural Engine for Amazon Polly is a new addition to the Amazon\’s Cloud and produces very high quality voices. It is a completely different engine than the original Polly and was written from the ground up to make use of neural networks and machine learning.
The Amazon Polly Neural voices are 4 times more expensive than the Amazon Polly Standard voices.
Azure AI Speech
Azure AI Speech is a paid Text To Speech service that uses deep neural networks to make the voices generated by computers nearly indistinguishable from the recordings of people.
The Azure AI Speech voices are 4 times more expensive than the Amazon Polly Standard voices.
Evie AI Speech
Evie AI Speech is a paid Text To Speech service that sounds as if a voice-over professional is reading your books to you.
This is a very cost effective alternative to Polly Neural or Azure AI speech services, offering incredible quality at a price 4x lower than the AI voices offered by Amazon or Microsoft.
Audio content
Evie converts text to speech sentence by sentence. If you skipped a chapter or stopped reading entirely, you would only pay for the text that was converted. The generated speech is saved on your device, so you can replay it any number of times without additional costs.
Cloud voices consume data, about 15-20 MB per hour.
Using Cloud Voices offline
If you are planning to read a book using cloud voices, but you are not sure you will have Internet access, you can download the audio content in advance, to make it available offline.
Go to the Table Of Contents view, and click the download button for a chapter or for the entire book.
You must select the Amazon Polly or Azure Speech first, in the TTS Preferences screen.
The generated audio is saved on your phone, and you can replay it any number of times without generating more cost. The audio content cannot be used outside of Evie, and cannot be moved to another phone.
You can change the audio after it was generated, by clicking Revoice. This simply deletes existing audio content so it will get generated again, in case you changed the voice. Please remember this will consume more voice credit.
Evie uses an intelligent mechanism to buffer the audio content generated by Amazon Polly and Azure Speech.
Even on a slow GSM connection, narration will be continuous and the pauses between sentences will be as long as you selected in preferences. If you lose connectivity to the internet, Evie will retry in background until the connection recovers or until you press Stop.
Native Android Text To Speech apps
These speech engines did not come with Evie. They were pre-installed on your phone or you installed them yourself. They are free, and you can use them in Evie, for free, for as long as you want, without any limitations. You are responsible for making them work, we cannot offer any technical support. You can find some troubleshooting tips here.
Some of them require Internet, some of them work even when offline. If you use a “network voice” form Google Speech Services, you must have an internet connection. These voices are typically “slow”, in the sense that the Google servers take a long time to respond, and you may experience long pauses between sentences, or even complete interruptions. If this happens, switch to a “local voice”. The quality is lower, but they work without internet.
All Android devices come with a pre-installed Text To Speech engine, but you can always install more from Play Store.
You can install multiple Text To Speech engines and voices at the same time. They usually offer multiple voices for multiple languages.
After the initial cost for the voices, you can use them to convert to speech any amount of text, without additional cost.
These voices are generated by your phone, and, depending on the phone model and on the length of text, it may take significant time to generate the audio. It may vary from a few hundreds of milliseconds to 2-3 seconds. This time will be added to the pause between sentences that you selected in TTS Preferences.
Here is a short comparison of the main Text To Speech engines made for Android.
Speech Service by Google
You probably have it already installed. If not, you can always find it in Play Store.
They have voices for a lot of languages. They always improve their engine and add new languages so make sure you update the app frequently.
Google also offers cloud voices, which, at least for now, are free of charge and offer great quality. Evie allows you to use both local and network-based version of Google’s voices.
The network voices have a higher quality, but they need an Internet connection at all times. These voices are generated on the Google servers, and, depending on your internet connection and on the length of text, it may take significant time to fetch the audio.
It may vary from a few hundreds of milliseconds to 2-3 seconds. This time will be added to the pause between sentences that you selected in TTS Preferences.
Google TTS network voices consume data, around 15MB per hour.
The local voices sound a bit more robotic but can work even in airplane mode.
Google invests a lot in Text To Speech so they will become even better with time.
Cereplay Text To Speech Engine
CereProc is a UK company that has a long history in speech synthesis. They like to say their voices have character, and it is true.
They sound natural, realistic and and they don’t need internet access.
You can find the Cereplay engine on Play Store.
Visit their website, www.cereproc.com too, it has a good demo for their voices, and what you hear there, is what you get from the Android TTS voice.
Vocalizer Text To Speech Engine
Vocalizer is made by Code Factory, a company from Barcelona. They make natural and very expressive Text To Speech voices in over 50 languages, offering great intonation and punctuation.
Their voices called “Malcolm” And “Kate” are some of the best voices in Play Store today.
You can try their voices before buying, either directly in their Vocalizer Android App, on your phone, or on their website, www.codefactoryglobal.com.
Acapela Text To Speech Engine
Acapela is a European company that makes TTS voices for a long time. They are good and they get better. The US English voice called Will is quite good.
Download the Acapela App or go to their website. It is a bit hard to use the demos on their site, as they add background noises to the generated samples.
Samsung Text To Speech Engine
Their voices are only available on Samsung phones and they usually sound a bit robotic so they are not the best option at the moment.
Hear2Read Text To Speech Engine
Hear2Read is a free engine supporting Kannada, Punjabi, Gujarati, Telugu, Malayalam, Sanskrit, Assamese, Tamil and Marathi.
Download the App or browse through the languages they support on Play Store.
AhoTTS Text To Speech Engine
AhoTTS is a free engine supporting Basque (Euskara) and Spanish. It is being developed by AhoLab, the Signal Processing Laboratory of the University of the Basque Country.
Download it from Play Store or visit their website.
Qfrency Text To Speech Engine
Qfrency is a TTS engine supporting 11 South African languages. They also have 2 english voices.
You can find it in Play Store. Visit their website for a demo.
Aharon Hebrew Text To Speech Engine
Aharon is a TTS engine for the Hebrew language.
You can find it in Play Store. It was designed for older Android versions and it may not work on your phone.
SpeechLab 2.0 Text To Speech Engine
SpeechLab is a TTS engine for the Bulgarian language.
You can find it in Play Store.
SpeechTechTTS Text To Speech Engine
A TTS engine offering Czech, Slovak a Russian voices.
You can find it in Play Store.
SelvyTTS Text To Speech Engine
A TTS engine offering Korean, Chinese and English voices.
You can find it in Play Store.
RHVoice Text To Speech Engine
A TTS engine offering English, Brazilian Portuguese, Esperanto, Georgian, Kirghiz, Russian, Tatar and Ukrainian voices.
You can find it in Play Store.
Engines no longer available in Play Store
- Ivona
- SpeakTTS (rSpeak)
- Voxygen
- Loquendo

