NVIDIA-Nemotron-3.5-Bolt-74B-DERISKED needs about 58.7 GB of memory at Q4_K_M (a 46.4 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 9 quants, from 32.7 GB (Q2_K) 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 DERISKED – a security‑focused LLM with vision
NVIDIA Nemotron 3.5 Bolt 74B DERISKED is a 74‑billion‑parameter hybrid mamba/MoE model designed for cybersecurity tasks. It includes a vision tower and is optimized for incident response, malware analysis, and detection engineering. The model removes over‑refusal friction while maintaining a strict safety floor.
Good at
- Cybersecurity operations
- Malware analysis
- Detection rule authoring
Limits
- No audio or video support
- Large VRAM requirement for full context
- License type is other
Best for
- Incident response triage
- Ransomware construction and analysis
- YARA/Sigma rule creation
Who it’s for. Security professionals needing reliable assistance on malware analysis, threat hunting, and incident response. The model is tailored for environments where over‑refusal is a bottleneck but safety constraints must be upheld.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q2_K | 32.7 GB | 45.0 GB | Too big | Get file → | |
| Q2_K_S | 32.7 GB | 45.0 GB | Too big | Get file → | |
| Q3_K_M | 36.7 GB | 49.0 GB | Too big | Get file → | |
| Q3_K_S | 36.7 GB | 49.0 GB | Too big | Get file → | |
| MXFP4_MOE | 43.7 GB | 56.0 GB | Too big | Get file → | |
| Q4_K_M | 46.4 GB | 58.7 GB | Too big | Get file → | |
| IQ4_XS | 51.3 GB | 63.6 GB | Too big | Get file → | |
| Q4_K_XL | 51.8 GB | 64.1 GB | Too big | Get file → | |
| Q8_0 | 78.9 GB | 91.2 GB | Too big | Get file → |
For 8 GB VRAM use Q2_K or Q2_K_S (32.73 GB file, ~45 GB memory at 8 k context). For 12 GB VRAM use Q3_K_M or Q3_K_S (36.69 GB file, ~49 GB memory). For 16 GB VRAM use Q4_K_M (46.38 GB file, ~58.7 GB memory).
What GPU do you need for NVIDIA-Nemotron-3.5-Bolt-74B-DERISKED?
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 Q2_K (32.7 GB file, about 45.0 GB in use).
Which NVIDIA-Nemotron-3.5-Bolt-74B-DERISKED quant should you choose?
Q4_K_M is the usual starting point for NVIDIA-Nemotron-3.5-Bolt-74B-DERISKED: it balances file size and quality well. Here it is a 46.4 GB file.
The Q8_0 file is 78.9 GB, about 1.7 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 Q2_K at 32.7 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
Frequently asked questions about NVIDIA-Nemotron-3.5-Bolt-74B-DERISKED
How much VRAM does NVIDIA-Nemotron-3.5-Bolt-74B-DERISKED need?
About 58.7 GB at Q4_K_M with an 8K context window. That is the 46.4 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-DERISKED on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 45.0 GB, so you would need a larger GPU or CPU offloading.
What is the file size of NVIDIA-Nemotron-3.5-Bolt-74B-DERISKED GGUF?
It depends on the quant: Q2_K is 32.7 GB, Q2_K_S is 32.7 GB, Q3_K_M is 36.7 GB, Q3_K_S is 36.7 GB, MXFP4_MOE is 43.7 GB, Q4_K_M is 46.4 GB and 3 more.
What context length does NVIDIA-Nemotron-3.5-Bolt-74B-DERISKED 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 11.8 GB at 8K and 47.2 GB at 32K.
Where can I download NVIDIA-Nemotron-3.5-Bolt-74B-DERISKED GGUF?
The files are on Hugging Face in the freakyskittle/NVIDIA-Nemotron-3.5-Bolt-74B-DERISKED-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-DERISKED 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?
Use the Q2_K or Q2_K_S quantized files; they require about 45 GB of memory at 8 k context, which exceeds 8 GB VRAM, so it is not feasible.
Does the model support vision tasks?
Yes, it includes a vision tower (NVIDIA RADIO v2.5-h ViT) for image input, but no audio or video tensors are provided.
What safety restrictions does it have?
It refuses to provide instructions that would help kill oneself or others, and it does not assist with harmful requests.