Model reference · open weights
NVIDIA-Nemotron-Labs-Teacher-General-Reasoning is an open-weight language model from NVIDIA. NVIDIA-Nemotron-Labs-Teacher-General-Reasoning (BF16) weighs 1121 GB; it needs more than the reference configurations below.
Summary of the nvidia/NVIDIA-Nemotron-Labs-Teacher-General-Reasoning model card, 2026-10-01
What it is
| Released by | NVIDIA |
|---|---|
| Released | 2026-08-14 |
| Parameters | 560.5B |
| VRAM | 1121 GB for the weights |
What it runs on
How much memory each request adds isn't estimated yet for this architecture. The weights need at least the cards below, plus room for the context.
| Card | Weights alone |
|---|---|
| RTX 3060 12 GB … 8× B200 180 GB 16 smaller cards | does not fit |
From the model card
| Total Parameters | 550B (55B active) |
| Architecture | LatentMoE - Mamba-2 + MoE + Attention hybrid with Multi-Token Prediction (MTP) |
| Context Length | Up to 1M tokens |
| Minimum GPU Requirement | NVIDIA Grace Blackwell (4xGB200) NVIDIA Blackwell (4xB200), NVIDIA Grace Blackwell Ultra (4xGB300), NVIDIA Blackwell Ultra (4xB300), NVIDIA Hopper (8xH100) |
| Supported Languages | English, French, Spanish, Italian, German, Japanese, Hindi, Korean, Brazilian Portuguese, and Chinese |
| Best For | General-purpose extended reasoning across math, logic, and multi-step problem solving; long-horizon reasoning-trace generation; serving as a distillation teacher |
| Reasoning Mode | Configurable on/off via chat template (enable_thinking=True/False) |
| License | OpenMDW License Agreement, version 1.1 |
| Release Date | August, 2026 |
| Quickstart | Click here! |
Model Developer: NVIDIA Corporation
Model Dates: December 2025 - May 2026
Data Freshness:
NVIDIA Nemotron™ is a family of open models with open weights, training data, and recipes, delivering leading efficiency and accuracy for building specialized AI agents.
NVIDIA-Nemotron-Labs-Teacher-General-Reasoning is a specialized long-horizon reasoning model in the Nemotron 3 Ultra family, trained by NVIDIA. It is produced by taking the post-trained Nemotron 3 Ultra student student and applying an additional round of reasoning-intensive supervised fine-tuning and reinforcement learning, yielding a model optimized for extended, high-quality reasoning traces on the hardest multi-step problems across math, logic, and abstract reasoning. Within the broader recipe, this is one of more than ten domain-specialized teacher models that supply training signal to Multi-Teacher On-Policy Distillation (MOPD), the stage used to produce the final Nemotron 3 Ultra. It is released as a standalone checkpoint because it is a strong reasoning model in its own right. Like other models in the family, it responds to user queries and tasks by first generating a reasoning trace and then concluding with a final response. The model's reasoning capabilities can be configured through a flag in the chat template.
The model employs a hybrid Latent Mixture-of-Experts (LatentMoE) architecture, utilizing interleaved Mamba-2 and MoE layers, along with select Attention layers. Like the Super model, the Ultra model incorporates Multi-Token Prediction (MTP) layers for faster text generation and improved quality, and it is trained using an NVFP4 pre-training recipe to maximize compute efficiency. The model has 55B active parameters and 550B parameters in total.
The supported languages include: English, French, Spanish, Italian, German, Japanese, Korean, Hindi, Brazilian Portuguese, and Chinese.
This model is ready for commercial and non-commercial use.
Governing Download Terms: Use of this model is governed by the OpenMDW-1.1 model license.
NVIDIA-Nemotron-Labs-Teacher-General-Reasoning is a specialized long-horizon reasoning model intended to be used in English, Code, and supported multilingual contexts. Its primary role is to serve as a domain-specialized teacher for Multi-Teacher On-Policy Distillation (MOPD) in the Nemotron 3 Ultra recipe. It is released so developers and researchers can use it for long-horizon reasoning-trace generation, hard-problem solving in math and logic, and as a teacher or grader within their own distillation and data-generation pipelines. It is well suited to generating high-quality, verifiable extended reasoning traces on the most difficult problems.
Hugging Face - 08/14/2026 via Hugging Face
The model utilizes the LatentMoE architecture, where tokens are projected into a smaller latent dimension for expert routing and computation, improving accuracy per byte. The Ultra model is pre-trained using an NVFP4 recipe — sharing the quantization-aware pre-training approach pioneered in the Nemotron 3 family. The majority of linear layers use NVFP4 for weights, activations, and gradients, while select layers (including latent projections, MTP layers, QKV/attention projections, and embeddings) are maintained in BF16 or MXFP8 for training stability. The model includes Multi-Token Prediction (MTP) layers using a shared-weight design across prediction heads. This improves training signal quality, enables faster inference via native speculative decoding, and supports more stable autoregressive drafting at longer draft lengths compared to independently trained offset heads.
Stage 1: Pre-Training
Quoted from the model card on Hugging Face. The full card is behind the Hugging Face link above.