Request for NVIDIA NIM API Rate Limit Increase (40 → 200 RPM) (original) (raw)

Hello NVIDIA Support Team,

I am writing to request a rate limit increase for my NVIDIA NIM API account to support an ongoing research and development workflow.

Account Details:

Rate Limit:

**Use Case:**I am developing a multi-agent research framework for quantitative finance experimentation. The system orchestrates several reasoning agents in parallel — each one decomposing a research question into multi-step tool-augmented chains (data retrieval, document parsing, structured extraction, and synthesis). Because a single research task expands into many sequential model calls across these chained agents, the current 40 RPM ceiling is frequently saturated mid-run, stalling experiments and making it difficult to evaluate end-to-end latency and throughput characteristics of the pipeline.

The 200 RPM tier would let me run the agent graph at its intended concurrency, complete multi-step research chains without throttling, and properly benchmark the framework’s performance. I am currently standardizing on Nemotron models via NIM as the backbone for this work and would like to continue building on the platform.

Thank you for your help!