Nemotron 3.5 Lightning
A hybrid NVIDIA model with an official NVFP4 checkpoint for agent inference.
- Published size
- 30B
- Active parameters
- 3B
- Context window
- 1M
- Architecture
- Hybrid MoE
- License
- OpenMDW 1.1
Deployment considerations
Stored precision and compute precision can differ. Check the official recipe for the exact GPU architecture and inference backend.
NVIDIA lists 30B total and 3B active parameters. This profile covers the official NVFP4 checkpoint released on August 11, 2026.
GPU memory estimates and minimum / recommended configurations are pending. This profile does not contain measured deployment results.
GPU requirements
- Minimum
- —GPU / VRAM · pending
- Recommended
- —GPU / VRAM · pending
Memory estimates and tested configurations will appear here after checkpoint review and deployment testing.
No invented leaderboards.
Latency, throughput, and cost per token will appear here after a reproducible run. Until then, this page helps you understand the model—not predict its performance.
Nemotron 3.5 Lightning deployment FAQ
How much GPU memory does Nemotron 3.5 Lightning need?
The full checkpoint weight footprint is pending review. Minimum and recommended GPU configurations will be added after testing; the model name or active parameter count alone is not a memory requirement.
Has BenchGrid benchmarked Nemotron 3.5 Lightning?
Not yet. This profile contains publisher specifications and calculated weight-memory estimates. We do not currently publish measured latency, throughput, or cost per token for this model.
Where do these specifications come from?
The specifications are based on the official NVIDIA model card linked on this page. Memory estimates use the stated total parameter count, including inactive experts for MoE models. Nominal model sizes are labeled with ~.
Explore your compute options.
Check available hardware, quotas, and current pricing with the provider. These links are not verified deployments or performance recommendations.
