Marco Nano Instruct needs about 6.8 GB of memory at Q4_K_M (a 5.4 GB file) with an 8K context window, so it fits on a 8 GB GPU. This repository has 12 quants, from 3.3 GB (Q2_K) to 16.0 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 3.8 GB.
Marco Nano Instruct – 8B sparse MoE instruct model
Marco Nano Instruct is an 8B parameter sparse Mixture‑of‑Experts model based on Qwen3‑moe architecture. It activates only 0.6B parameters per token and supports 32,768 context tokens. The model is licensed under Apache‑2.0.
Good at
- Reasoning
- Tool use
- Multilingual instruction
Limits
- Requires 8,000‑token context for full performance
- No vision support
- Large VRAM needed for higher‑quality quants
Best for
- Chatting in multiple languages
- Answering technical queries
- Generating code snippets
Who it’s for. Developers and researchers who need a lightweight, multilingual instruct model with strong reasoning and tool‑use capabilities. The model is suitable for edge deployments where memory is limited but high‑quality instruction following is required.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q2_K | 3.3 GB | 4.8 GB | Fits | Get file → | |
| Q3_K_S | 3.9 GB | 5.3 GB | Fits | Get file → | |
| Q3_K_M | 4.2 GB | 5.6 GB | Fits | Get file → | |
| Q3_K_L | 4.4 GB | 5.8 GB | Fits | Get file → | |
| IQ4_XS | 4.5 GB | 5.9 GB | Fits | Get file → | |
| Q4_K_S | 4.9 GB | 6.3 GB | Fits | Get file → | |
| Q4_K_M | 5.4 GB | 6.8 GB | Fits | Get file → | |
| Q5_K_S | 5.7 GB | 7.1 GB | Fits | Get file → | |
| Q5_K_M | 6.1 GB | 7.6 GB | Tight | Get file → | |
| Q6_K | 7.2 GB | 8.6 GB | Too big | Get file → | |
| Q8_0 | 8.5 GB | 10.0 GB | Too big | Get file → | |
| F16 | 16.0 GB | 17.5 GB | Too big | Get file → |
For 8 GB VRAM, use Q2_K (file size 3.34 GB, estimated 4.8 GB at 8k context). For 12 GB VRAM, Q4_K_S (4.91 GB, 6.3 GB). For 16 GB VRAM, Q5_K_M (6.11 GB, 7.6 GB). Higher‑quality quants such as Q8_0 or F16 require 10 GB+ VRAM.
What GPU do you need for Marco Nano Instruct?
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: the best quant that fits is IQ4_XS (4.5 GB file, about 5.9 GB in use).
- 8 GB VRAM: the best quant that fits is Q5_K_M (6.1 GB file, about 7.6 GB in use).
- 12 GB VRAM: the best quant that fits is Q8_0 (8.5 GB file, about 10.0 GB in use).
- 16 GB VRAM: the best quant that fits is Q8_0 (8.5 GB file, about 10.0 GB in use).
- 24 GB VRAM: the best quant that fits is F16 (16.0 GB file, about 17.5 GB in use).
Which Marco Nano Instruct quant should you choose?
Q4_K_M is the usual starting point for Marco Nano Instruct: it balances file size and quality well. Here it is a 5.4 GB file.
The Q8_0 file is 8.5 GB, about 1.6 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 3.3 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
Source
Frequently asked questions about Marco Nano Instruct
How much VRAM does Marco Nano Instruct need?
About 6.8 GB at Q4_K_M with an 8K context window. That is the 5.4 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 Marco Nano Instruct on an 8 GB GPU?
Yes, with Q5_K_M: the 6.1 GB file needs about 7.6 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Marco Nano Instruct GGUF?
It depends on the quant: Q2_K is 3.3 GB, Q3_K_S is 3.9 GB, Q3_K_M is 4.2 GB, Q3_K_L is 4.4 GB, IQ4_XS is 4.5 GB, Q4_K_S is 4.9 GB and 6 more.
What context length does Marco Nano Instruct support?
The GGUF header lists a trained context length of 32,768 tokens (32K). The KV cache grows with the context you use: about 0.9 GB at 8K and 3.8 GB at 32K.
Where can I download Marco Nano Instruct GGUF?
The files are on Hugging Face in the mradermacher/Marco-Nano-Instruct-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 Marco Nano Instruct 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 Marco Nano Instruct on 8 GB VRAM?
Yes, the Q2_K quant uses about 4.8 GB of VRAM at 8k context, fitting within an 8 GB GPU.
Does the model support vision inputs?
No, the model only processes text and does not include vision capabilities.
What license applies to this model?
The model is released under the Apache‑2.0 license.