NVIDIANemotron 3
Nemotron 3 Nano 4B
Nano 4B is a compact text model compressed from Nemotron Nano 9B v2 with NVIDIA's Elastic method.
What the release provides
The release has about 3.97B parameters. Its hybrid architecture mainly uses Mamba-2 and MLP layers with four attention layers. The card lists context up to 262K and system-prompt control over reasoning.Nemotron 3 Nano 4B official model card
How to use it well
Practical guidance. Use it for a narrow local task with a clearly defined output. Compare direct responses with reasoning-enabled responses on difficult examples. Long context still consumes resources, so retain only the evidence that the current step needs.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | Downloadable |
| Model identifiers | nvidia/NVIDIA-Nemotron-3-Nano-4B-BF16 |
| Release license | NVIDIA Nemotron Open Model LicenseNemotron 3 Nano 4B official model card |