NVIDIA: Nemotron 3 Nano 30B A3B (free)
nvidia
NVIDIA Nemotron 3 Nano 30B A3B is a small language MoE model with highest compute efficiency and accuracy for developers to build specialized agentic AI systems. The model is fully...
- Context window
- 256,000 tokens
- Input cost
- $0.05 / 1M
- Output cost
- $0.20 / 1M
- Latency (p50)
- —