Hi NVIDIA NIM Team,
I’m writing to kindly request an increase in my free-tier API rate limit.
Account Email: rihansaifi4849@gmail.com
Current Limit: 40 RPM
Requested Limit: 200 RPM (or the highest tier available for individual developers)
Use Case:
I’m conducting applied research on multi-step agentic workflows for code analysis. My architecture chains 4–6 inference calls per user prompt (planning → tool selection → execution → verification).
With the 40 RPM limit, a single agent loop triggers 429 errors within seconds, breaking the flow and making iterative testing impossible.
I specifically rely on Qwen for structured reasoning and Step-3.5 Flash for speed. Both are excellent for my use case, but the shared 40 RPM pool across endpoints becomes the bottleneck.
Increasing the limit to 200 RPM would reduce context-switching overhead by ~70% and allow me to validate agent behaviors without artificial throttling.
Thank you for supporting open research with NIM.
Best regards,
rayn