Uploaded July 2026 | Updated November 2026, 7 hours ago
Turn text into speech using an ESP32 and an online API. We'll look at bot wit.ai and OpenAI TTS. Give your project a voice!
Article with code: dronebotworkshop.com/text-to-speech
More articles and tutorials: dronebotworkshop.com
Join the conversation on the forum: forum.dronebotworkshop.com
Subscribe to the newsletter and stay in touch: dronebotworkshop.com/subscribe
Text-to-speech systems are not exactly new; the first experiments date back to the late 1930’s, and we have had appliances that “talk” for years. But up until about 10 years ago, the technology produced speech that was fairly “robotic” (think Stephen Hawking) or limited to specific words or phrases.
All that changed with the introduction of Neural Network TTS systems, artificial intelligence that knows how to speak. And today, we are going to experiment with Neural Network TTS systems using an ESP32.
Instead of trying to cram all of this onto an ESP32, which isn’t really capable of handling the task, we will make use of a couple of online services that expose APIs to convert incoming text into speech.
We will be working with both wit.ai (a product of Meta) and OpenAI TTS. We’ll see how to use these API services with an ESP32, and, of course, we will test them and compare voice quality.
Here is the Table of Contents for today's video:
00:00 - Introduction
01:48 - Text-To-Speech Systems
06:49 - wit.ai
18:47 - OpenAI TTS
25:59 - Conclusion
Both of these services have their advantages and disadvantages, allowing you to choose between price (wit.ai is free at the level we are using it; OpenAI TTS incurs a very small charge) and performance (OpenAI TTS sounds more natural than wit.ai). Either of them can be used to add voice output capability to your next ESP32 project.
I hope you enjoy the video!
Bill
Turn text into speech using an ESP32 and an online API. We'll look at bot wit.ai and OpenAI TTS. Give your project a voice!
Article with code: dronebotworkshop.com/text-to-speech
More articles and tutorials: dronebotworkshop.com
Join the conversation on the forum: forum.dronebotworkshop.com
Subscribe to the newsletter and stay in touch: dronebotworkshop.com/subscribe
Text-to-speech systems are not exactly new; the first experiments date back to the late 1930’s, and we have had appliances that “talk” for years. But up until about 10 years ago, the technology produced speech that was fairly “robotic” (think Stephen Hawking) or limited to specific words or phrases.
All that changed with the introduction of Neural Network TTS systems, artificial intelligence that knows how to speak. And today, we are going to experiment with Neural Network TTS systems using an ESP32.
Instead of trying to cram all of this onto an ESP32, which isn’t really capable of handling the task, we will make use of a couple of online services that expose APIs to convert incoming text into speech.
We will be working with both wit.ai (a product of Meta) and OpenAI TTS. We’ll see how to use these API services with an ESP32, and, of course, we will test them and compare voice quality.
Here is the Table of Contents for today's video:
00:00 - Introduction
01:48 - Text-To-Speech Systems
06:49 - wit.ai
18:47 - OpenAI TTS
25:59 - Conclusion
Both of these services have their advantages and disadvantages, allowing you to choose between price (wit.ai is free at the level we are using it; OpenAI TTS incurs a very small charge) and performance (OpenAI TTS sounds more natural than wit.ai). Either of them can be used to add voice output capability to your next ESP32 project.
I hope you enjoy the video!
Bill










