Fastino Nemotron 3.5 Lightning Finance needs about 26.4 GB of memory at IQ4_NL (a 18.9 GB file) with an 8K context window, so it fits on a 32 GB GPU. Longer context needs more: at 32K tokens the KV cache alone adds about 27.9 GB.
Fastino Nemotron 3.5 Lightning Finance – 31.6B parameter financial reasoning model
Fastino Nemotron 3.5 Lightning Finance is a 31.6B parameter mixture‑of‑experts model specialized for financial document reasoning, numeric question answering, and source‑grounded research. It is fine‑tuned with LoRA on a large financial dataset and released under Apache 2.0.
Good at
- Financial reasoning and numeric QA
- Table and SEC disclosure extraction
- Source‑grounded financial research
Limits
- No vision support
- Requires large VRAM (≈26.4 GB at 8 k context)
- Licensed under Apache 2.0
Best for
- Financial analysts needing precise numeric extraction
- Researchers building finance‑specific LLM applications
- Developers integrating finance reasoning into chatbots
Who it’s for. This model is intended for developers and researchers who need a specialized financial reasoning engine. It is suitable for teams with access to high‑memory GPUs or multi‑GPU setups.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| IQ4_NL | 18.9 GB | 26.4 GB | Too big | Get file → |
The only available quantization is IQ4_NL. The 18.9 GB file requires about 26.4 GB of VRAM at 8 k context. For 32 GB or larger GPUs, this quant is suitable. For smaller GPUs, the model is not supported.
What GPU do you need for Fastino Nemotron 3.5 Lightning Finance?
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: the best quant that fits is IQ4_NL (18.9 GB file, about 26.4 GB in use).
Which Fastino Nemotron 3.5 Lightning Finance quant should you choose?
IQ4_NL (18.9 GB) is the middle option in this repository. The usual sweet spot between file size and quality, and the quant most people start with.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Frequently asked questions about Fastino Nemotron 3.5 Lightning Finance
How much VRAM does Fastino Nemotron 3.5 Lightning Finance need?
About 26.4 GB at IQ4_NL with an 8K context window. That is the 18.9 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 Fastino Nemotron 3.5 Lightning Finance on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 26.4 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Fastino Nemotron 3.5 Lightning Finance GGUF?
It depends on the quant: IQ4_NL is 18.9 GB.
What context length does Fastino Nemotron 3.5 Lightning Finance 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 7.0 GB at 8K and 27.9 GB at 32K.
Where can I download Fastino Nemotron 3.5 Lightning Finance GGUF?
The files are on Hugging Face in the hemono/Fastino-Nemotron-3.5-Lightning-Finance-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 Fastino Nemotron 3.5 Lightning Finance use?
Hugging Face lists the license as apache-2.0. 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 Fastino Nemotron 3.5 Lightning Finance on 8 GB VRAM?
No, the model requires approximately 26.4 GB of VRAM at 8 k context.
Is there a smaller quantization available?
Only the IQ4_NL quantization is provided in the repository.
Does the model support vision modalities?
No, the model supports only text modalities.