NVIDIA Nemotron Nano 9B v2 needs about 16.4 GB of memory at Q4_K_M (a 6.5 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 26 quants, from 0.0 GB to 17.8 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 37.6 GB.
NVIDIA Nemotron Nano 9B v2 – 9B parameter LLM with reasoning control
NVIDIA Nemotron Nano 9B v2 is a 9‑billion parameter language model trained from scratch by NVIDIA. It uses a hybrid Mamba‑2 and MLP architecture with four attention layers. The model supports reasoning traces that can be enabled or disabled via a system prompt.
Good at
- Reasoning with controllable thinking budget
- Multilingual support (English, German, Spanish, French, Italian, Japanese)
- General purpose chat and coding tasks
Limits
- Requires large VRAM for full precision (≈27.7 GB at 8 k context)
- No vision or audio modalities
- License governed by NVIDIA Open Model License Agreement
Best for
- Developers building reasoning‑aware chatbots
- Researchers testing hybrid Mamba‑Transformer models
- Applications needing configurable reasoning depth
Who it’s for. Engineers and researchers who need a general‑purpose reasoning LLM with configurable thinking budget. The model is suitable for commercial use under NVIDIA’s license.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.0 GB | 9.9 GB | Too big | Get file → | |
| IQ2_S | 5.0 GB | 14.9 GB | Too big | Get file → | |
| IQ2_M | 5.0 GB | 14.9 GB | Too big | Get file → | |
| Q2_K | 5.0 GB | 14.9 GB | Too big | Get file → | |
| IQ3_XXS | 5.1 GB | 15.0 GB | Too big | Get file → | |
| IQ3_XS | 5.1 GB | 15.0 GB | Too big | Get file → | |
| Q3_K_S | 5.1 GB | 15.0 GB | Too big | Get file → | |
| IQ3_M | 5.2 GB | 15.1 GB | Too big | Get file → | |
| IQ4_XS | 5.3 GB | 15.2 GB | Too big | Get file → | |
| Q2_K_L | 5.3 GB | 15.2 GB | Too big | Get file → | |
| IQ4_NL | 5.3 GB | 15.2 GB | Too big | Get file → | |
| Q4_0 | 5.3 GB | 15.2 GB | Too big | Get file → | |
| Q3_K_M | 5.4 GB | 15.3 GB | Too big | Get file → | |
| Q3_K_L | 5.5 GB | 15.4 GB | Too big | Get file → | |
| Q3_K_XL | 5.8 GB | 15.7 GB | Too big | Get file → | |
| Q4_1 | 5.8 GB | 15.7 GB | Too big | Get file → | |
| Q4_K_S | 6.2 GB | 16.1 GB | Too big | Get file → | |
| Q4_K_M | 6.5 GB | 16.4 GB | Too big | Get file → | |
| Q4_K_L | 6.7 GB | 16.6 GB | Too big | Get file → | |
| Q5_K_S | 6.8 GB | 16.7 GB | Too big | Get file → | |
| Q5_K_M | 7.1 GB | 17.0 GB | Too big | Get file → | |
| Q5_K_L | 7.3 GB | 17.1 GB | Too big | Get file → | |
| Q6_K | 9.1 GB | 19.0 GB | Too big | Get file → | |
| Q6_K_L | 9.1 GB | 19.0 GB | Too big | Get file → | |
| Q8_0 | 9.5 GB | 19.4 GB | Too big | Get file → | |
| BF16 | 17.8 GB | 27.7 GB | Too big | Get file → |
For 8 GB VRAM, use IQ2_S or IQ2_M (≈5 GB file, ≈14.9 GB memory at 8 k context). For 12 GB VRAM, Q3_K_S or IQ3_XS (≈5.13 GB file, ≈15 GB memory). For 24 GB VRAM, Q4_K_M or Q5_K_M (≈6.53 GB or 7.07 GB file, ≈16.4–17 GB memory).
What GPU do you need for NVIDIA Nemotron Nano 9B v2?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: no quant fits at an 8K context.
- 6 GB VRAM: no quant fits at an 8K context.
- 8 GB VRAM: no quant fits at an 8K context.
- 12 GB VRAM: the best quant that fits is the main file (0.0 GB file, about 9.9 GB in use).
- 16 GB VRAM: the best quant that fits is Q4_1 (5.8 GB file, about 15.7 GB in use).
- 24 GB VRAM: the best quant that fits is Q8_0 (9.5 GB file, about 19.4 GB in use).
- 32 GB VRAM: the best quant that fits is BF16 (17.8 GB file, about 27.7 GB in use).
Which NVIDIA Nemotron Nano 9B v2 quant should you choose?
Q4_K_M is the usual starting point for NVIDIA Nemotron Nano 9B v2: it balances file size and quality well. Here it is a 6.5 GB file.
The Q8_0 file is 9.5 GB, about 1.4 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is the first one in the table at 0.0 GB, with the quality trade-offs that come with fewer bits.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about NVIDIA Nemotron Nano 9B v2
How much VRAM does NVIDIA Nemotron Nano 9B v2 need?
About 16.4 GB at Q4_K_M with an 8K context window. That is the 6.5 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run NVIDIA Nemotron Nano 9B v2 on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 9.9 GB, so you would need a larger GPU or CPU offloading.
What is the file size of NVIDIA Nemotron Nano 9B v2 GGUF?
It depends on the quant: file is 0.0 GB, IQ2_S is 5.0 GB, IQ2_M is 5.0 GB, Q2_K is 5.0 GB, IQ3_XXS is 5.1 GB, IQ3_XS is 5.1 GB and 20 more.
What context length does NVIDIA Nemotron Nano 9B v2 support?
The GGUF header lists a trained context length of 1,048,576 tokens (1024K). The KV cache grows with the context you use: about 9.4 GB at 8K and 37.6 GB at 32K.
Where can I download NVIDIA Nemotron Nano 9B v2 GGUF?
The files are on Hugging Face in the bartowski/nvidia_NVIDIA-Nemotron-Nano-9B-v2-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does NVIDIA Nemotron Nano 9B v2 use?
Hugging Face lists the license as other. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run NVIDIA Nemotron Nano 9B v2 on 8 GB VRAM?
Use a low‑bit quant like IQ2_S or IQ2_M; the model file is ~5 GB and requires ~15 GB VRAM at 8 k context, so 8 GB is insufficient without further optimization.
Does the model support vision or audio inputs?
No, the model only handles text inputs and outputs.
What license applies to this model?
It is governed by the NVIDIA Open Model License Agreement, allowing commercial use under its terms.