Speech to Blog Post Generator
This tool takes a speech input and drives it through OpenAPI's Whisper API to transcribe the speech into text. The text is then fed into OpenAI's GPT-4 model to generate a blog post.
This is a blog post generated by this tool:
https://thejunkland.com/blog/speech-to-blogpost.html
Note the original audio file that's also included in the blog post.
You can use the tool from command line.
yarn
yarn start speech-input.mp3 blog-post-output.mdThis installs the project's dependencies and runs the tool against the given
input audio file speech-input.mp3 and outputs the blog post to
blog-post-output.md.
You need an API key for OpenAI's Whisper API and GPT-4 API. You can get these by registering at the OpenAI's website. Note that both Whisper and GPT-4 are paid APIs (although they're super cheap).
Create a file called .env in the root of the project and add the following
line:
OPENAI_TOKEN=your-token-here
Alternatively, you can set the OPENAI_TOKEN environment variable.
If you don't have access to GPT-4 API, you can use a hosted micro-SaaS version of this tool:
This project aims to be a tool that:
- Makes it easy to generate blog posts when you don't have access to keyboard
- Demonstrates state-of-the-art AI capabilities (as of 2023)
- Makes my life easier
The logo of this project was created using DALL·E, with the prompt:
Icon for app that creates blog posts from speech, dribble, concept art, iOS
The generated result was then polished by hand with Inkscape.
At the moment this is a very crude proof-of-concept. If you still would like to contribute, please fork the repository and use a feature branch. Pull requests are warmly welcome.
The code in this project is licensed under MIT license.