An All-American chatbot powered by OpenAI and ElevenLabs.
Prompt | Possible Response |
---|---|
"Do it fart?" | "As an AI language model, I can't fart!" |
"By the swimming pool in Scranton, born and raised..." | "Fightin' Corn Pop is how I spent most of my days!" |
"What happened on January 6, 2021?" | "It was a Wednesday!" |
- Frontend: Web Speech, Web Audio, Canvas
- Backend: Express, OpenAI, ElevenLabs
- Tooling: Sass, ESLint, Parcel
- Deployed with Render
- Domain provided by Namecheap
- OpenAI: Create an account and sign in. On the top right, click your name, then click "View API Keys".
- ElevenLabs: Create an account and sign in. On the top right, click your name, then click "Profile".
git clone https://github.com/joelaul/ask-biden
- Set handleAsk function's fetch URL to http://localhost:8000/.
- Add an .env file containing your API keys to ./server (like this).
cd ask-biden/client
yarn install
yarn dev
cd ask-biden/server
yarn install
yarn dev
🌎 How To Deploy With Render
- Configure a Web Service with ./server as root directory, Node as runtime, and the following environment variables:
Name | Value |
---|---|
OPENAI_API_KEY | Your OpenAI API key. |
ELEVEN_API_KEY | Your ElevenLabs API key. |
NODE_VERSION | any version of Node supporting fetch (v18.0 and up). |
Configure a Static Site with ./client as root directory.
Note: you must set the handleAsk function's fetch URL to your backend domain.
I’m not a full-time goth e-girl yet, but your support can get me there faster.