This is a little application that will allow users to pay XXX amount of Satoshis for a XXXms session with ChatGPT-4. You can test this application live here

There are bound to be bugs and performance issues, but we wanted to release this for anyone who wanted to learn more about lightning, chatGPT and making cutting edge webapps.

This software is open source and free to all. ( See LICENSE.md )

On page-load the user is assigned a UUID kept in local.storage ( So they'd lose access if they cleared their browser cache ). This UUID is what allows us to track whether a user has paid.

The flow operation:

1. We create an invoice using the /lightning/createInvoice route
2. User pays the invoice
3. We call /lightning/checkInvoice route - if invoice was paid, then set a KV of {uuid: Date.now()}
4. We can now call /chat route ( checks that their uuid:timestamp is within the duration )
5. We return the GPT access

• We probably could have used a better session validation solution ( not storing a UUID in local.storage ) - however it was easy and quick to implement. So we stuck with the KISS principle.
• Our site is using https://legend.lnbits.com/ - which is fine for testing but not production. I believe they shut off access every once and a while so people don't rely on them
• We initially did pay-per-question, but the user experience was shit

If you want to run and deploy this yourself, you'll have to, clone the repo, and fill in the .env file

You'll have to setup a Vercel KV project: https://vercel.com/docs/storage/vercel-kv/quickstart

First clone the Repo git clone https://github.com/project-excalibur/Lightning-GPT.git

Then copy .env.example -> .env

Then you'll have to fill in all of the appropriate fields in the .env file

NEXT_PUBLIC_API_URL=http://localhost:3000/api - This is how the app knows where to route the api calls. When hosting you will have to replace this with your url + /api

NEXT_PUBLIC_SESSION_TIME=300000 - This is how long the ChatGPT session lasts for in ms.

NEXT_PUBLIC_SESSION_COST=100 - How many sats per session

GPT_API_KEY=[YOUR_KEY_HERE] - This is your chat-gpt api key ( do not commit this to github ). You can get one here

LNBITS_API_KEY=[YOUR_KEY_HERE] - This is what allows you to interface with the lightning network, we suggest you spin one up with Ledgend LNBits here. Do not use this for any production application

KV_REST_API_READ_ONLY_TOKEN=[YOUR_KEY_HERE]
KV_REST_API_TOKEN=[YOUR_KEY_HERE]
KV_REST_API_URL=[YOUR_URL_HERE]
KV_URL=[YOUR_URL_HERE]

These are your Vercel VK variables. Most of this is auto-generated using vercel env pull .env.development.local which will output to the file .env.development.local

Lastly run pnpm install pnpm dev