Account Details:
NVIDIA Account Email: ennui1020@gmail.com
API Key ID (Last 4 chars): 8w5-
Request for Rate Limit Increase
Model: DeepSeek-V4-Pro (nvidia_nim/deepseek-ai/deepseek-v4-pro)
Current Rate Limit: 40 RPM
Use Case:
I am using this model with Hermes Agent (an open-source autonomous AI agent framework) for research and development work. The agent performs complex multi-step reasoning with tool calling, which generates multiple API calls in short bursts.
Expected Usage:
Approximately 100-200 RPM during active sessions
Research and personal AI assistant development
Non-commercial use
Current Issues:
Frequent 429 errors even with client-side throttling at 18 RPM using LiteLLM proxy.
I would greatly appreciate if you could increase the limit to 150 RPM or higher.
Thank you!