nvidia
NVIDIA-Nemotron-Nano-9B-v2-Japanese
- Parameters
- 8.9B
- Max context length
- 1,048,576
- Architecture
- nemotron_h
- License
- Commercial OK
- Base model
- nvidia/NVIDIA-Nemotron-Nano-9B-v2-Japanese
- License
- nvidia-open-model-license
- llama.cpp
- b6315+
About NVIDIA-Nemotron-Nano-9B-v2-Japanese
NVIDIA-Nemotron-Nano-9B-v2-Japanese (nvidia) is a GGUF model based on nvidia/NVIDIA-Nemotron-Nano-9B-v2-Japanese on Hugging Face. GGUF files are published by 4 sources (mmnga-o/NVIDIA-Nemotron-Nano-9B-v2-Japanese-gguf, hiratagoh/NVIDIA-Nemotron-Nano-9B-v2-Japanese-GGUF, DevQuasar/nvidia.NVIDIA-Nemotron-Nano-9B-v2-Japanese-GGUF and others). It has about 8.9B parameters, the nemotron_h architecture, a maximum context length of 1,048,576 tokens (values written in the GGUF).
There are 20 quantizations, with file sizes from 5.01 GB (Q2_K) to 17.8 GB (F16) (actual sizes on Hugging Face).
Running Q4_K_M (6.53 GB) with an 8,192-token context needs an estimated 7.61 GB of memory, and an estimated 8.01 GB at 32,768 tokens. For 8,192 tokens, a GPU with 8 GB or more of VRAM (such as GeForce RTX 3070 Ti, Arc A580), or a Mac with 8 GB or more of unified memory, is a rough guide (macOS caps how much memory the GPU can use, so leave some headroom). The smallest, Q2_K (5.01 GB), needs an estimated 6.09 GB at 8,192 tokens.
The original model is licensed under NVIDIA Nemotron Open Model License, which allows commercial use (keep the license text and copyright notice when redistributing). Check the original license for the terms of use.
File size and estimated memory per quantization
- Up to 8 GB
- Up to 16 GB
- Up to 24 GB
- Up to 48 GB
- Up to 128 GB
- More
File sizes are the actual values on Hugging Face. Memory is estimated for each context length (KV cache f16). The formula is in How memory is calculated.
All memory figures are Estimated (KV cache f16). It should run on a setup with at least this much VRAM (or Mac unified memory). Click a value to see the devices and models that fit that amount (Find by GPU / Mac).
4 sources publish GGUF files of this model. Quantization types and file sizes differ slightly between sources, so there is a table for each source below.
About the "quality guide" (how far a quantization drifts from the original)
For each quantization type, how far the output drifts from the original model, grouped into six levels from values llama.cpp measured on Llama-3-8B.
The value is mean KLD (the difference between the original and quantized model in next-token probabilities); smaller means closer to the original. Quantizations made with an imatrix (importance data) come out slightly smaller. Levels use the value without an imatrix (or with one, for types that only have that).
These values were not measured on this model. Larger models and MoE models tend to drift less with the same type. Distributor-specific mixed types (such as UD-Q4_K_XL) cannot be matched by type name alone, so they show "-".
| Quantization | Quality guide | KLD (no imatrix) | KLD (imatrix) | Size (Llama-3-8B) |
|---|---|---|---|---|
| Q8_0 | Nearly identical | 0.0014 | - | 7.96 GiB |
| Q6_K | Very small | 0.0055 | - | 6.14 GiB |
| Q5_K_M | Very small | 0.0108 | - | 5.33 GiB |
| Q5_K_S | Small | 0.0166 | - | 5.21 GiB |
| Q5_1 | Small | 0.0180 | - | 5.65 GiB |
| Q5_0 | Small | 0.0222 | - | 5.21 GiB |
| Q4_K_M | Small | 0.0313 | 0.0282 | 4.58 GiB |
| Q4_K_S | Medium | 0.0431 | 0.0320 | 4.37 GiB |
| IQ4_NL | Small | - | 0.0357 | 4.35 GiB |
| IQ4_XS | Small | - | 0.0363 | 4.14 GiB |
| Q4_1 | Medium | 0.0717 | - | 4.78 GiB |
| Q4_0 | Medium | 0.0719 | - | 4.34 GiB |
| Q3_K_L | Medium | 0.0872 | 0.0731 | 4.03 GiB |
| Q3_K_M | Medium | 0.1019 | 0.0844 | 3.74 GiB |
| IQ3_M | Medium | - | 0.1025 | 3.53 GiB |
| IQ3_S | Medium | - | 0.1113 | 3.42 GiB |
| IQ3_XS | Large | - | 0.1387 | 3.28 GiB |
| IQ3_XXS | Large | - | 0.1836 | 3.05 GiB |
| Q3_K_S | Large | 0.2282 | 0.1998 | 3.41 GiB |
| IQ2_M | Large | - | 0.3260 | 2.74 GiB |
| Q2_K | Very large | 0.4451 | 0.3322 | 2.96 GiB |
| Q2_K_S | Very large | - | 0.4034 | 2.96 GiB |
| IQ2_S | Very large | - | 0.4392 | 2.56 GiB |
| IQ2_XS | Very large | - | 0.5463 | 2.43 GiB |
| IQ2_XXS | Very large | - | 0.8120 | 2.24 GiB |
| IQ1_M | Very large | - | 1.3931 | 2.01 GiB |
| IQ1_S | Very large | - | 2.2546 | 1.88 GiB |
Source: llama.cpp (MIT), measured on Llama-3-8B against FP16, Wikitext-2, revision f364eb6f. Original table · Reading and choosing quantizations
GGUF source
mmnga-o/NVIDIA-Nemotron-Nano-9B-v2-Japanese-gguf
Converted and published by a third party About this label
Last checked: Nov 11, 2026
- Architecture
- nemotron_h
- llama.cpp b6315 and later
- Parameters
- 8.9B
- Max context length
- 1,048,576
- License
- nvidia-nemotron-open-model-license Commercial OK
Ollama ollama run hf.co/mmnga-o/NVIDIA-Nemotron-Nano-9B-v2-Japanese-gguf:Q4_K_M
llama.cpp llama-server -hf mmnga-o/NVIDIA-Nemotron-Nano-9B-v2-Japanese-gguf:Q4_K_M
GGUF source
hiratagoh/NVIDIA-Nemotron-Nano-9B-v2-Japanese-GGUF
Converted and published by a third party About this label
Last checked: Nov 11, 2026
- Architecture
- nemotron_h
- llama.cpp b6315 and later
- Parameters
- 8.9B
- Max context length
- 1,048,576
- License
- nvidia-nemotron-open-model-license Commercial OK
| Quantization | File size Measured | Quality guide | 4,096 | 8,192 | 16,384 | 32,768 | 65,536 | 131,072 | 262,144 | 1,048,576 max |
|---|---|---|---|---|---|---|---|---|---|---|
| IQ4_XS | 5.27 GB | Small | 6.29 GB | 6.35 GB | 6.49 GB | 6.75 GB | 7.29 GB | 8.37 GB | 10.5 GB | 23.4 GB |
| IQ4_NL | 5.31 GB | Small | 6.33 GB | 6.39 GB | 6.53 GB | 6.80 GB | 7.33 GB | 8.41 GB | 10.6 GB | 23.4 GB |
| Q4_K_M | 6.53 GB | Small | 7.54 GB | 7.61 GB | 7.74 GB | 8.01 GB | 8.55 GB | 9.62 GB | 11.8 GB | 24.7 GB |
| Q5_K_M | 7.07 GB | Very small | 8.09 GB | 8.15 GB | 8.29 GB | 8.56 GB | 9.09 GB | 10.2 GB | 12.3 GB | 25.2 GB |
| Q6_K | 9.14 GB | Very small | 10.2 GB | 10.2 GB | 10.4 GB | 10.6 GB | 11.2 GB | 12.2 GB | 14.4 GB | 27.3 GB |
| Q8_0 | 9.46 GB | Nearly identical | 10.5 GB | 10.5 GB | 10.7 GB | 10.9 GB | 11.5 GB | 12.6 GB | 14.7 GB | 27.6 GB |
| BF16 | 17.8 GB | Original precision | 18.8 GB | 18.9 GB | 19.0 GB | 19.3 GB | 19.8 GB | 20.9 GB | 23.0 GB | 35.9 GB |
Ollama ollama run hf.co/hiratagoh/NVIDIA-Nemotron-Nano-9B-v2-Japanese-GGUF:Q4_K_M
llama.cpp llama-server -hf hiratagoh/NVIDIA-Nemotron-Nano-9B-v2-Japanese-GGUF:Q4_K_M
GGUF source
DevQuasar/nvidia.NVIDIA-Nemotron-Nano-9B-v2-Japanese-GGUF
Converted and published by a third party About this label
Last checked: Nov 11, 2026
- Architecture
- nemotron_h
- llama.cpp b6315 and later
- Parameters
- 8.9B
- Max context length
- 1,048,576
- License
- nvidia-nemotron-open-model-license Commercial OK
Ollama ollama run hf.co/DevQuasar/nvidia.NVIDIA-Nemotron-Nano-9B-v2-Japanese-GGUF:Q4_K_M
llama.cpp llama-server -hf DevQuasar/nvidia.NVIDIA-Nemotron-Nano-9B-v2-Japanese-GGUF:Q4_K_M
GGUF source
tatsuyaaaaaaa/NVIDIA-Nemotron-Nano-9B-v2-Japanese-gguf
Converted and published by a third party About this label
Last checked: Nov 11, 2026
- Architecture
- nemotron_h
- llama.cpp b6315 and later
- Parameters
- 8.9B
- Max context length
- 1,048,576
- License
- nvidia-nemotron-open-model-license Commercial OK
Ollama ollama run hf.co/tatsuyaaaaaaa/NVIDIA-Nemotron-Nano-9B-v2-Japanese-gguf:Q4_K_M
llama.cpp llama-server -hf tatsuyaaaaaaa/NVIDIA-Nemotron-Nano-9B-v2-Japanese-gguf:Q4_K_M
Generation speed on common GPUs and Macs (estimated)
A rough guide to how fast Q4_K_M (6.53 GB) generates text. Each token reads the model weights once, so the upper limit is memory bandwidth ÷ file size. The guide is 60–80% of that, where most published measurements fall.
| Device | Memory bandwidth | Guide (tokens/s) Estimated | Upper limit |
|---|---|---|---|
| Arc A550M 8 GB | 224 GB/s | 21–27 | 34 |
| Radeon RX 6800 16 GB | 512 GB/s | 47–63 | 78 |
| GeForce RTX 5090 32 GB | 1,792 GB/s | 165–220 | 275 |
| Apple M2 16 GB | 100 GB/s | 9.2–12 | 15 |
| Apple M2 Pro 16 GB | 200 GB/s | 18–25 | 31 |
| Apple M5 Ultra 96 GB | 1,200 GB/s | 110–147 | 184 |
Estimates only. Prompt processing speed (limited by compute) is not included. Longer contexts and offloading part of the model to the CPU make it slower. Macs are shown only when installed memory is at least 4/3 of what is needed. Devices without published bandwidth (some GeForce cards, etc.) are not shown.
License and commercial use
NVIDIA Nemotron Open Model License Commercial OK
Commercial use, modification and redistribution are allowed (NVIDIA states the works are commercially usable). The grant is irrevocable. When redistributing, include the license text, and if a NOTICE file exists, add "Licensed by NVIDIA Corporation under the NVIDIA Nemotron Model License." in derivatives. If you sue over patent or copyright infringement by the model, your license for it ends.
Based on the license on the original model's Hugging Face page. This differs from the GGUF repository's label (nvidia-open-model-license). Quantized files are made from the original model, so check the original model's license terms as well. Original license text (checked Nov 1, 2026)
The summary states facts this site read from the original text and is not legal advice. Always check the original license and its latest terms before use. All licenses
How to run
Mainline llama.cpp can load this model's format (nemotron_h) from b6315 (Aug 28, 2025) onward. Ollama and LM Studio bundle llama.cpp internally, so an older app fails to load it with errors such as "unknown model architecture". If that happens, update the app.
Example commands using Q4_K_M. To use another quantization, replace the name with one from the table above. Commands for files from other sources are in each source's section. How to run local LLMs
Ollama
Run it by pointing directly at the GGUF on Hugging Face.
ollama run hf.co/mmnga-o/NVIDIA-Nemotron-Nano-9B-v2-Japanese-gguf:Q4_K_M
llama.cpp
Pass the repository and quantization name to llama-server (or llama-cli) with -hf.
llama-server -hf mmnga-o/NVIDIA-Nemotron-Nano-9B-v2-Japanese-gguf:Q4_K_M
LM Studio
Paste mmnga-o/NVIDIA-Nemotron-Nano-9B-v2-Japanese-gguf into the search box, download it, and choose the quantization from the file list.
Other models from nvidia
-
7.5B
nvidia
NVIDIA-Nemotron-3.5-Bolt-E4B
- UD-IQ2_M size
- 3.53 GB
- Memory at 8K
- 4.78 GB
Measured Estimated VRAM 6 GB+
-
31.6B
nvidia
NVIDIA-Nemotron-3.5-Lightning-30B-A3B
- Q4_0 size
- 18.9 GB
- Memory at 8K
- 19.8 GB
Measured Estimated VRAM 24 GB+
-
7.5B
nvidia
NVIDIA-Nemotron-3.5-Bolt-E4B-4GB
- Q4_K_M size
- 6.18 GB
- Memory at 8K
- 7.43 GB
Measured Estimated VRAM 8 GB+
-
8B
nvidia
Nemotron-3-Embed-8B
- Q4_K_M size
- 4.90 GB
- Memory at 8K
- 6.84 GB
Measured Estimated VRAM 8 GB+
-
32.9B
nvidia
NVIDIA-Nemotron-3.5-Transcribe-30B-A4B
- F16 size
- 1.67 GB
- Memory at 8K
- 2.58 GB
Measured Estimated VRAM 4 GB+
-
nvidia
NVIDIA-Nemotron-3.5-Super-VL-120B-A12B
- NVFP4 size
- 1.35 GB
- Memory at 8K
- 2.40 GB
Measured Estimated VRAM 4 GB+
Models of a similar size (other developers)
-
9.2B
ornith-ai
Ornith-1.5-9B
- Q4_K_M size
- 5.78 GB
- Memory at 8K
- 7.17 GB
Measured Estimated VRAM 8 GB+
-
11.9B
google
gemma-4-12B-it
- Q4_K_M size
- 7.12 GB
- Memory at 8K
- 8.83 GB
Measured Estimated VRAM 12 GB+
-
9B
ornith-ai
Ornith-1.0-9B
- Q4_K_M size
- 5.63 GB
- Memory at 8K
- 7.02 GB
Measured Estimated VRAM 8 GB+
-
9B
DavidAU
Qwen3.5-9B-The-Defiant-Fable-Uncensored-Heretic-NEO-IMATRIX-MAX-MTP
- Q4_K_M size
- 6.83 GB
- Memory at 8K
- 8.20 GB
Measured Estimated VRAM 8 GB+
-
OpenMOSS-Team
MOSS-Music-8B-Instruct
- F16 size
- 1.72 GB
- Memory at 8K
- 3.78 GB
Measured Estimated VRAM 4 GB+
-
8.2B
Qwen
Qwen3-VL-8B-Instruct
- Q4_K_M size
- 5.03 GB
- Memory at 8K
- 7.08 GB
Measured Estimated VRAM 8 GB+
Frequently asked questions
- How much memory do I need to run NVIDIA-Nemotron-Nano-9B-v2-Japanese (Q4_K_M)?
- The file size is 6.53 GB (measured). With a 8,192-token context, the estimated memory including the KV cache is 7.61 GB. The formula is published at https://localllm.az-arte.com/en/guide/memory-calculation.
- Can NVIDIA-Nemotron-Nano-9B-v2-Japanese be used commercially?
- The original model's license is NVIDIA Nemotron Open Model License (Commercial use allowed).Commercial use, modification and redistribution are allowed (NVIDIA states the works are commercially usable). The grant is irrevocable. When redistributing, include the license text, and if a NOTICE file exists, add "Licensed by NVIDIA Corporation under the NVIDIA Nemotron Model License." in derivatives. If you sue over patent or copyright infringement by the model, your license for it ends. Check the original license text before deciding.
- Does NVIDIA-Nemotron-Nano-9B-v2-Japanese run in 8 GB, 12 GB, 16 GB or 24 GB of VRAM?
- 8 GB: up to Q5_K_M (est. 8.15 GB); 12 GB: up to Q8_0 (est. 10.5 GB); 16 GB: up to Q8_0 (est. 10.5 GB); 24 GB: up to Q8_0 (est. 10.5 GB) (estimated at 8,192 tokens with an f16 KV cache, files from the most-downloaded source).
- What is the command to run NVIDIA-Nemotron-Nano-9B-v2-Japanese with Ollama?
- Run "ollama run hf.co/mmnga-o/NVIDIA-Nemotron-Nano-9B-v2-Japanese-gguf:Q4_K_M" to use the GGUF on Hugging Face directly. To use another quantization, replace the name at the end.