NE

NVIDIA-Nemotron-3.5-Bolt-74B-A9B

NemotronGGUF by AtomicChatUpdated 2026-10-28
Download on Hugging Face
Parameters74B
Context256K
Layers88
Architecturenemotron_h_moe
Smallest quant21.9 GB
Quick answer

NVIDIA-Nemotron-3.5-Bolt-74B-A9B needs about 64.0 GB of memory at Q4_K_M (a 51.7 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 13 quants, from 21.9 GB (IQ1_M) to 78.9 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 47.2 GB.

NVIDIA Nemotron 3.5 Bolt 74B‑A9B – a large MoE model for local customization

NVIDIA Nemotron 3.5 Bolt 74B‑A9B is a 74‑billion‑parameter mixture‑of‑experts model with 9‑billion active parameters. It is designed for post‑training, domain adaptation, and quantized deployment on local hardware. The model is released under the OpenMDW License Agreement v1.1.

Good at

  • Post‑training and fine‑tuning
  • Domain adaptation
  • Quantization for local inference

Limits

  • Requires at least 21.94 GB VRAM for the smallest quant
  • No vision support in this GGUF
  • Context length not specified

Best for

  • Custom AI agents
  • Research evaluation
  • Domain‑specific fine‑tuning

Who it’s for. Developers and researchers who need a large MoE model for local customization and post‑training. The model is suitable for those who want to adapt the model to specific tasks or domains and produce quantized variants for deployment.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ1_M21.9 GB 34.3 GB
Too big Get file →
IQ1_M22.5 GB 34.8 GB
Too big Get file →
IQ2_XXS23.0 GB 35.3 GB
Too big Get file →
IQ2_XXS26.7 GB 39.0 GB
Too big Get file →
IQ2_XS29.1 GB 41.4 GB
Too big Get file →
IQ2_M31.5 GB 43.8 GB
Too big Get file →
IQ3_XXS36.3 GB 48.7 GB
Too big Get file →
IQ3_M40.2 GB 52.5 GB
Too big Get file →
IQ4_XS46.6 GB 58.9 GB
Too big Get file →
Q4_K_M51.7 GB (2 parts) 64.0 GB
Too big Get file →
Q5_K_M61.1 GB (2 parts) 73.4 GB
Too big Get file →
Q6_K72.5 GB (2 parts) 84.8 GB
Too big Get file →
Q8_078.9 GB (2 parts) 91.2 GB
Too big Get file →

For a 24 GB VRAM budget use IQ1_M (21.94 GB, 34.3 GB memory at 8 k context). For 32 GB VRAM use IQ2_XXS (22.96 GB, 35.3 GB memory). For 48 GB VRAM use IQ3_XXS (36.34 GB, 48.7 GB memory). For 64 GB VRAM use Q4_K_M (51.71 GB, 64 GB memory). The memory estimates are approximate.

What GPU do you need for NVIDIA-Nemotron-3.5-Bolt-74B-A9B?

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: no quant fits at an 8K context.
  • 16 GB VRAM: no quant fits at an 8K context.
  • 24 GB VRAM: no quant fits at an 8K context.
  • 32 GB VRAM: no quant fits at an 8K context.
  • 48 GB VRAM: the best quant that fits is IQ2_M (31.5 GB file, about 43.8 GB in use).

Which NVIDIA-Nemotron-3.5-Bolt-74B-A9B quant should you choose?

Q4_K_M is the usual starting point for NVIDIA-Nemotron-3.5-Bolt-74B-A9B: it balances file size and quality well. Here it is a 51.7 GB file.

The Q8_0 file is 78.9 GB, about 1.5 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 IQ1_M at 21.9 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturenemotron_h_moe
Parameters74B (approximate, mixture of experts)
Active parameters9B
Layers88
Trained context262,144 tokens
KV cache (fp16)1408 KB per token

Source

Quantized byAtomicChat
FamilyNemotron
Licenseother
Downloads (HF)18,961
Last updated2026-10-28

Frequently asked questions about NVIDIA-Nemotron-3.5-Bolt-74B-A9B

How much VRAM does NVIDIA-Nemotron-3.5-Bolt-74B-A9B need?

About 64.0 GB at Q4_K_M with an 8K context window. That is the 51.7 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-3.5-Bolt-74B-A9B on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 34.3 GB, so you would need a larger GPU or CPU offloading.

What is the file size of NVIDIA-Nemotron-3.5-Bolt-74B-A9B GGUF?

It depends on the quant: IQ1_M is 21.9 GB, IQ1_M is 22.5 GB, IQ2_XXS is 23.0 GB, IQ2_XXS is 26.7 GB, IQ2_XS is 29.1 GB, IQ2_M is 31.5 GB and 7 more.

What context length does NVIDIA-Nemotron-3.5-Bolt-74B-A9B support?

The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 11.8 GB at 8K and 47.2 GB at 32K.

Where can I download NVIDIA-Nemotron-3.5-Bolt-74B-A9B GGUF?

The files are on Hugging Face in the AtomicChat/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-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-3.5-Bolt-74B-A9B 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 this model on 8 GB VRAM?

No, the smallest quant requires 21.94 GB VRAM.

Does this model support vision tasks?

No, this GGUF release does not include vision support.

What license governs this model?

The model is governed by the OpenMDW License Agreement, version 1.1.

Other GGUFs of NVIDIA-Nemotron-3.5-Bolt-74B-A9B-BF16

Similar models

Browse all →

More from AtomicChat

All by AtomicChat →

More Nemotron models

All Nemotron →

Other large models

All large models →