rlimit.com is doing great, open beta launch soon 🙌
<100ms usually, no cache involved yet, but you are right, there will be an option to enable optimistic responses. Meaning if there is plenty tokens in the bucket for the key left, it will just cache it for a while and do updates in the background 🙌 does it make sense?
https://rlimit.com prelaunch
- Rate-limiting made easy; literally, fetch a URL
- Just works with no libraries, and scales well
- Observability out of the box
- Pay for allowed requests only
Would you use it? Try it now, and sign up for launch notification
Made a ChatGPT plugin template with authentication using Clerk, ready to build new plugins 😎
https://github.com/eidam/chatgpt-plugin-clerk-auth
If you read this you are probably not rate limited meaning you are in a good place 🙌
You need to make it slower/unusable so you can make it fast and call it a feature (not a PM either)
Same, love it
If it’s just a couple of days in year you just somewhat survive through it I guess 😅
Heyyy 🙌
Thank you, invited new folks 🙏
I made a thing (yet another AI assistant) with @openai.bsky.social function calling and the new @vercel.com AI SDK. 🔥
Check it out superhelp.ai 👈
👋 hello world