Hi, What managed platforms for developing and hosting ChatGPT plugins are you using? I’ve seen a few people using https://replit.com/, but what else? Also, pros and cons for your choice will be much appreciated.
Thanks!
Hi, What managed platforms for developing and hosting ChatGPT plugins are you using? I’ve seen a few people using https://replit.com/, but what else? Also, pros and cons for your choice will be much appreciated.
Thanks!
I have used hetzner for everything for the past few years, it has been great!
Pros:
Very professional, very quick service, good connectivity and zero downtime
Cons:
Not easy for first timers and costumer support beyond a set amount will cost you money.
You can also ask GPT about “Using AWS Lambda with Amazon API Gateway for REST API”.
Github Codespaces and CodeSandbox are also mentioned in the documentation but it is really up to you and what you feel comfortable with. Digital ocean, heroku, firebase, vercel, AWS, Azure or GCP. Whatever that can host your API.
The ChatGPT-specific requirements are essentially two static files. When it comes to hosting, that decision should be dictated by the needs of the API being provided to ChatGPT through the instructions in those two files.
For prototyping, localhost will provide the tightest loop for iterations. And the OpenAI plugin quick start repo will get you started in no time: GitHub - openai/plugins-quickstart: Get a ChatGPT plugin up and running in under 5 minutes!
All these are great suggestions, thanks! I was thinking of something more suitable for non-pros that includes low-code and fully-managed services. The way I see it it shouldn’t be much different than developing a standard API, still as mentioned by @wfhbrian it really depends on the use case. I’m trying to figure out if there are any general groups that plugin use cases can be divided and if there are any specific requirements development and hosting platforms should meet. Do you think there such separation can be defined or it’s too early?
Replit is free and has a built in Ai code system. It’s my favourite.
What are the scalability limitations? Is the free tier only suitable for hobby or also for professional usage?
I use it for all my stuff on Reddit. I had several million views of my bots last month and never paid for any hosting
I’m using gcp cloud run because I’m familiar with it and it integrates with the rest of my stack.
I don’t think you need that much horsepower.
Any of the services mentioned above will work just fine.
https://replit.com/ is a great starting point. If it begins to falter, https://railway.app/ does a better job at scaling. Meanwhile you can also get a Cloudflare load balancer for your API endpoint and scale horizontally on the backend (I recall @ruv wrote about this somewhere).
If all these are still not enough, you are probably one of the most popular plugins on ChatGPT. At that point it’ll make sense to build your own infra on AWS (or perhaps Azure? since ChatGPT is on Azure)
How are you using it so that its free? How do you keep your plugins continuously running without the always on feature?