Nemotron 3.5 Super — reviews, specs & pricing
NVIDIA's open hybrid Mamba-Transformer MoE model, 120B parameters with 12B active, for reasoning, long context and agentic workloads.
Summary
Open 120B (12B active) NVIDIA model for reasoning and agents.
Sample use case
Reasoning, long-context and agentic workloads
Specifications
- Provider: nvidia
- License: open-weight
- Parameters: 120B
- Released: 2026-10-14
Average rating 0.0 from 0 community reviews on Reviuws.