Using gTTS in Node-RED with python-venv node #45
404background
started this conversation in
Flow & Ideas
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Also introduced in my article below:
In English: https://404background.com/en/programming/en-python-gtts/
In Japanese: https://404background.com/program/python-gtts/
In this post, I tried using gTTS (Google Text-to-Speech) with the python-venv node in Node-RED.
Using gTTS, you can generate an audio file that speaks text.
Just follow the steps below to make it available.
The following flow can be used to execute gTTS.

Here is the flow:
If you pass only text, for example, you can generate an audio file of the text entered at the dashboard node.


I used template nodes to pass the arguments of the executable.
In addition, connecting to a node that plays audio, the generated audio file can be played.
Here is the flow:
Beta Was this translation helpful? Give feedback.
All reactions