NVIDIANemotron 3.5
Nemotron 3.5 Bolt 74B-A9B
Bolt 74B-A9B is a compressed vision-language derivative of Nemotron 3.5 Super VL. It has 74B total and 9B active parameters.
What the release provides
The release uses native 2-bit quantization for a 24GB deployment target. It supports text, images and video with context up to 256K. The architecture search budget used batch size one and 128K context.Nemotron 3.5 Bolt 74B-A9B official model card
How to use it well
Practical guidance. A memory target is tied to its stated configuration. Recheck memory with your image count, context and concurrent sessions. For browser actions, review the proposed target and take a new screenshot after each action to confirm the resulting state.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | Downloadable |
| Model identifiers | nvidia/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB |
| Release license | OpenMDW 1.1Nemotron 3.5 Bolt 74B-A9B official model card |