← open-weight

Llama 3.1 Nemotron 70B

NVIDIA · USA
NVIDIA's Llama 3.1 fine-tune with custom RLHF. Inherits Llama 3.1 Community License terms. Strong conversational quality; useful default when you want Llama behaviour with NVIDIA's alignment.
Licence facts
Licence
Llama 3.1 Community + NVIDIA Community Model terms
Commercial use
Below 700M MAU
Derivatives
Allowed with naming rule
Attribution
'Built with Llama' required
Parameters
70B dense
Training data
Base Llama + NVIDIA SteerLM / HelpSteer2
Last updated
2024-Q4
Known risks
  • Two stacked licences (Meta + NVIDIA) — verify both
  • MAU cap
  • Alignment dataset (HelpSteer2) is disclosed but small
Sources
→ Model card→ Llama licence
See also
NVIDIA NIM· APILlama 3.3 70B· open-weight
Reviewed by Ali Madjaji · Last reviewed 2026-04-15
Home·Methodology·About·/