NVIDIA-Nemotron-Nano-12B-v2
NVIDIA-Nemotron-Nano-12B-v2 is an open-weights language model from NVIDIA, released August 18, 2025. It has 12B parameters (dense).
NVIDIA-Nemotron-Nano-12B-v2 is a large language model (LLM) trained from scratch by NVIDIA, and designed as a unified model for both reasoning and non-reasoning tasks. It responds to user queries and tasks by first generating a reasoning trace and then concluding with a final response.
Check it against your own PC →
Can you run NVIDIA-Nemotron-Nano-12B-v2 locally?
The smallest common GPU it runs on is 6 GB of VRAM (with 32 GB of RAM): IQ4_XS, split GPU / CPU, roughly 6 tokens per second. Estimates assume 32 GB of DDR5 RAM, a 16K context and llama.cpp-style GGUF quantization. Enter your exact hardware in the live app.
| GPU memory | Verdict | Best quant | VRAM | RAM | tok/s |
|---|---|---|---|---|---|
| 6 GB VRAM | Runs, slowly | IQ4_XS (good) | 5.2 GB | 4.4 GB | ~6 |
| 8 GB VRAM | Runs, slowly | Q5_K_M (excellent) | 7.2 GB | 4.6 GB | ~5 |
| 12 GB VRAM | Runs on GPU | Q4_K_M (very good) | 10.6 GB | – | ~28 |
| 16 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 13.2 GB | – | ~29 |
| 24 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 13.2 GB | – | ~53 |
| 32 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 13.2 GB | – | ~96 |
Ready-made quantizations on Hugging Face: QuantFactory/NVIDIA-Nemotron-Nano-12B-v2-GGUF · Mungert/NVIDIA-Nemotron-Nano-12B-v2-GGUF · DevQuasar/nvidia.NVIDIA-Nemotron-Nano-12B-v2-Base-GGUF
Benchmark results
Epoch AI has not published results for NVIDIA-Nemotron-Nano-12B-v2 yet. New models are usually evaluated within days to weeks.
Links
Announcement · Hugging Face: nvidia/NVIDIA-Nemotron-Nano-12B-v2-Base
More from NVIDIA
NVIDIA Nemotron Labs Dory1.0 4B BF16 · NVIDIA Nemotron Labs Dory1.0 4B MXFP8 · Nemotron 3.5 Lightning · Nemotron 3 Ultra · Nemotron 3 Super · Nemotron 3-Nano-30B-A3B · Nemotron-Cascade 14B · NVIDIA Isaac GR00T N1.6