pyttsx3 is a cross-platform, fully offline text-to-speech library for Python 3. Unlike cloud speech services, it converts text to spoken audio entirely on the local machine using the speech engines already present in the operating system, so it works with no internet connection and no API keys. The project has earned more than 2,500 stars on GitHub and was still actively updated in October 2026.
pyttsx3 wraps the native speech synthesizers of each platform behind one simple Python API: SAPI5 on Windows, NSSpeechSynthesizer on macOS, and eSpeak on Linux. A few lines of code are enough to make an application talk — initialize the engine, pass a string, and run the speaking loop. Developers can adjust the voice, speaking rate, and volume, enumerate the voices installed on the system, and either play speech through the speakers or save it to an audio file such as MP3 or WAV. Because everything happens locally, there are no per-character fees, no latency from network calls, and no privacy concerns about sending text to a third-party service.
pyttsx3 is free and open source under the MPL-2.0 license. There are no usage limits or paid plans; the only requirement is a speech engine available on the host operating system.
pyttsx3 suits Python developers who need speech output in desktop apps, accessibility tools, notification systems, kiosks, or offline and embedded projects. It is a practical default choice when you want a program to talk without depending on an internet connection or a paid speech API.