The Google text-to-speech API for voice is now open to developers for a small monthly fee to customize for different experiences…
The familiar voice which speaks from Google Maps and the Google Assistant is about to change. The search giant has announced it’s now opening the Google text-to-speech API to developers.
The Google Text-to-Speech Technology is Now Open to Developers
For a minimal monthly fee, devs can have access to Standard, (Non-WaveNet) voices, as well as WaveNet voices. The technology is one of the more advanced in its field. Here’s how the company describes it on the official page:
“Google Cloud Text-to-Speech enables developers to synthesize natural-sounding speech with 30 voices, available in multiple languages and variants. It applies DeepMind’s groundbreaking research in WaveNet and Google’s powerful neural networks to deliver the highest fidelity possible.”
Google is making the technology available so developers can apply it in different environments. For instance, inside call centers, IoT devices, and other formats. Currently, there are 32 different voices, available in 12 languages. Users are able to customize things like pitch, pace of speed, and volume.
Google used its DeepMind crew to improve text-to-speech. Instead of stringing speech fragments together to form words, it employs WaveNet technology. The end result is a less robotic sound and more natural speech. It’s also 1,000 times faster and can generate higher quality audio. Tests reveals listeners think WaveNet voices are 20 percent better than other types of voices. The technology reduces the quality gap between human speech and tech-generated speech by nearly 70 percent.