Open doors to more players with text to speech for games

Yes! It’s true that most TTS solutions for video games are cloud-based, streaming tools, so they require an internet connection. But ReadSpeaker' gaming plugins are different.

Our TTS engine runs on the gamer’s device, which eliminates the need for an internet connection. It also speeds up performance considerably.

Of course, we also offer more conventional streaming TTS through an API, if that’s what you’re looking for. Just choose the speechEngine SDK if so.

The answer depends on a number of factors, most notably the TTS voices you choose. Our top-quality AI voices—built on neural Hidden Markov Models (HMMs)—only occupy between 13 and 16 MB per voice.

Our DNN-based AI voices might take up to 32 MB per voice, however. We also use a technology called DNN Micro, which can run as low as 4 MB per voice.

This question also depends on the CPU your gamer’s device has available. If it supports up to 800 MHz or higher, it can run our DNN, HMM, and DNN Micro voices.

If the CPU only offers 400 MHz, it might be limited to our Unit Selection Synthesis (USS) voices, which take between 80 and 600 MB per voice. They still deliver high quality, just not as high as our best AI voices.

To sum it all up, the TTS footprint is based on the TTS voice, and can run anywhere from 4 to 600 MB per voice.

We’ll work with you to find the technology that offers the best combination of quality and footprint for your unique project.

Our gaming plugins provide TTS in more than 35 languages and dialects, including:

• Arabic
• Chinese (Cantonese)
• Chinese (Mandarin)
• Chinese (Taiwanese Mandarin)
• Czech
• Danish
• Dutch (Belgian)
• Dutch (Netherlander)
• English (American)
• English (Australian)
• English (British)
• English (Indian)
• English (Irish)
• English (New Zealander)
• English (Scottish)
• English (South African)
• Finnish
• French (Belgian)
• French (Canadian)
• French (French)
• German
• Hebrew
• Hindi
• Hungarian
• Indonesian
• Italian
• Japanese
• Korean
• Latvian
• Norwegian (Bokmål)
• Polish
• Portuguese (Brazilian)
• Portuguese (Europoean)
• Romanian
• Russian
• Slovak
• Spanish (Castilian)
• Spanish (Mexican)
• Spanish (Agentinian)
• Swedish
• Thai
• Turkish
• Vietnamese
• Welsh

Listen to voice samples of these languages (and more) here.

Don’t see what you’re looking for Contact us to discuss your needs. We’re always working on new TTS languages, and maybe yours is already in development! Reach out for a demo or voice samples, or to discuss your gaming use case.

ReadSpeaker can build an ethical custom voice (aka ethical voice clone) for you, but we are a B2B company, and we do not offer DIY or self-service voice cloning.

Ethical AI-powered Voice solutions are our core priority. We simply can’t guarantee ethical usage if we make our technology available to all.

When we create TTS voices, we set up fair contractual agreements across the TTS chain, from the voice actors to the user. That’s the only way to avoid exploitation or unauthorized use of voice data—and we don’t see a way to offer voice cloning ethically on the self-service model.

Learn more about our take on ethical AI Voice here.

The audio format breaks down accessibility barriers for lots of gamers, including players with blindness, low vision, dyslexia, attention deficits, cognitive disabilities, or a simple preference for auditory experience over reading.

The key to gaming accessibility is to offer each player lots of choice. Let them decide how they interact with your experience.

Every gamer brings a unique combination of needs, capabilities, and predispositions to your game. You want to meet each player where they are, and you can only do that with lots of options.

Some gamers may require audible UI narration to play, while others may do fine with text resizing or no accommodations at all.

But if you don’t offer TTS accessibility tools, you run the risk of leaving people behind—and shrinking your target audience considerably.

Although Gen AI speech and text-to-speech voices have advancements in the application of artificial intelligence as a common denominator, the latter typically involves generating voices from scratch, often requiring significant online processing. ReadSpeaker's AI-powered TTS voices, on the other hand, are built on deep learning models trained on high-quality recordings.

This ensures natural, expressive speech optimized for gaming, with a focus on performance and seamless integration. Crucially, our voices run offline directly within your game engine, giving you greater control and reducing latency to near zero. Essentially, ReadSpeaker's TTS leverages AI to deliver high-quality voice content with greater efficiency and flexibility for game developers.