Nemotron-4-340B
by NVIDIA · free · Last verified 2026-03-17
Nemotron-4-340B is NVIDIA's open-weight 340-billion-parameter model family, featuring both a base model and instruction-tuned variant optimized for synthetic data generation and RLHF pipelines. It delivers frontier performance on reasoning and knowledge benchmarks and is particularly valued as a teacher model for training smaller models.
https://huggingface.co/nvidia/Nemotron-4-340B-Instruct ↗C+
C+—Average
Adoption: C+Quality: AFreshness: ACitations: BEngagement: F
Specifications
- License
- NVIDIA Open Model License
- Pricing
- free
- Capabilities
- text-generation, reasoning, synthetic-data-generation, instruction-following, reward-modeling
- Integrations
- NVIDIA NIM, Hugging Face, Together AI
- Use Cases
- synthetic-data-generation, model-training, research, enterprise-inference
- API Available
- Yes
- Parameters
- 340B
- Context Window
- 4K
- Modalities
- text
- Training Cutoff
- 2023
- Tags
- nvidia, open-source, frontier, synthetic-data, reward-model
- Added
- 2026-03-17
- Completeness
- 100%
Index Score
54.7Adoption
55
Quality
86
Freshness
84
Citations
62
Engagement
0