Llama 3.1 Nemotron 51B Instruct
NVIDIA
Key facts · verified Oct 29, 2026
Llama 3.1 Nemotron 51B Instruct is a NVIDIA model released on 23 Sept 2024. Pricing: Open weights — free to self-host. It ranks #301 on LMArena's text leaderboard (Elo 1287, read 1 Nov 2026). Source: huggingface.co.
NVIDIA's Llama 3.1 Nemotron 51B Instruct, a 51-billion-parameter model derived from Meta's Llama 3.1 70B and published with open weights under the NVIDIA Open Model License.
Pricing & specs checkedSource: vendor documentation · checked Oct 29, 2026LMArena: 1287 Elo (rank 301) · source · retrieved Nov 1, 2026
Full Specifications
Open weights — free to self-host | |
|---|---|
| Identity | |
| Developer | NVIDIA |
| Released | Sep 2024 |
| Status | Open weights |
| Licence | NVIDIA Open Model License |
| Self-hostable | Yes — fits on a single NVIDIA H100; packaged as an NVIDIA NIM microservice |
| Cost | |
| Input price | Self-hosted — no first-party API price |
| Output price | Self-hosted — no first-party API price |
| Capability | |
| Vision in | No — text in, text out |
| Access | |
| Fine-tuning | Yes — open weights permit it |
| Measured quality | |
| LMArena Elo | 1287 (checked Nov 2026) |
| LMArena rank | Rank 301 |