Hello NVIDIA Team,
I would like to request an increase in my NVIDIA NIM API rate limit.
Current Limit: 40 RPM
Requested Limit: 200 RPM (or highest available tier)
Use Case:
AI agent development
multi-agent workflows
model testing and evaluation
bug bounty / research workflows
Reason:
My system generates multiple chained API calls per task, and I frequently hit the 40 RPM limit, resulting in 429 errors. This significantly slows development and testing.
Usage Type:
Non-production
Personal research and development
I am part of the Google x NVIDIA developer community and actively working on AI systems.
I will use the API responsibly with proper rate limiting and backoff.
Thank you for your support!
Just buy a cloud, you must be fair with other free tier user
Thank you for your comment. My request is specifically about the individual developer rate limit tier provided by NVIDIA, which is distinct from cloud compute resources. I’m working on multi-agent AI pipelines that require several sequential API calls. A modest increase would significantly help my development workflow without impacting other free-tier users. I’ve implemented client-side backoff and rate limiting.
I look forward to an official response from the NVIDIA team regarding any available options. Many thanks.
I know you are using AI to answer me, so let just simplify this, we are just free tier user normal like you the key of ts API is for user can try and small-dev production, we dont use AI for framework or multi-agent
I know you’ve struggling with 429 error, and me too but just endure, we cant do anything, models are overload, server are slow, queue are long, many ppl are using nim so let just think abt evryone not just you
I get where you’re coming from, but I’m well aware of the paid options. There’s no harm in asking if NVIDIA has any flexibility for researchers or specific developer programs. I’ll just wait for an official response from the team. Thanks.
ok feel free to do, bcause im aware of the serious status of overload model in nim