NinjaMouse 3B 40L Danube needs about 3.1 GB of memory at Q4_K_M (a 1.8 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 24 quants, from 0.7 GB (IQ1_S) to 2.4 GB (Q6_K).
NinjaMouse 3B 40L Danube – a 2.94B‑parameter coding‑focused Llama model
NinjaMouse 3B 40L Danube is a 2.94B‑parameter Llama model fine‑tuned on coding datasets. It supports a 16384‑token context window and is released under the apache‑2.0 license. The model is available in multiple GGUF quantizations.
Good at
- Code generation
- Debugging assistance
- Script writing
Limits
- No vision support
- Limited to 16384‑token context
- Requires agreement to license terms
Best for
- Developers needing lightweight code assistants
- Educational coding tools
- Rapid prototyping of scripts
Who it’s for. This model is intended for developers, educators, and hobbyists who need a small, efficient LLM for coding tasks. It is suitable for local deployment on GPUs with moderate VRAM.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| IQ1_S | 0.7 GB | 2.0 GB | Fits | Get file → | |
| IQ1_M | 0.7 GB | 2.1 GB | Fits | Get file → | |
| IQ2_XXS | 0.8 GB | 2.2 GB | Fits | Get file → | |
| IQ2_XS | 0.9 GB | 2.2 GB | Fits | Get file → | |
| IQ2_S | 1.0 GB | 2.3 GB | Fits | Get file → | |
| IQ2_M | 1.0 GB | 2.4 GB | Fits | Get file → | |
| Q2_K_S | 1.0 GB | 2.4 GB | Fits | Get file → | |
| Q2_K | 1.1 GB | 2.5 GB | Fits | Get file → | |
| IQ3_XXS | 1.2 GB | 2.5 GB | Fits | Get file → | |
| IQ3_XS | 1.2 GB | 2.6 GB | Fits | Get file → | |
| Q3_K_S | 1.3 GB | 2.6 GB | Fits | Get file → | |
| IQ3_S | 1.3 GB | 2.6 GB | Fits | Get file → | |
| IQ3_M | 1.4 GB | 2.7 GB | Fits | Get file → | |
| Q3_K_M | 1.4 GB | 2.8 GB | Fits | Get file → | |
| Q3_K_L | 1.6 GB | 2.9 GB | Fits | Get file → | |
| IQ4_XS | 1.6 GB | 2.9 GB | Fits | Get file → | |
| Q4_0 | 1.7 GB | 3.0 GB | Fits | Get file → | |
| IQ4_NL | 1.7 GB | 3.0 GB | Fits | Get file → | |
| Q4_K_S | 1.7 GB | 3.0 GB | Fits | Get file → | |
| Q4_K_M | 1.8 GB | 3.1 GB | Fits | Get file → | |
| Q4_1 | 1.9 GB | 3.2 GB | Fits | Get file → | |
| Q5_K_S | 2.0 GB | 3.4 GB | Fits | Get file → | |
| Q5_K_M | 2.1 GB | 3.4 GB | Fits | Get file → | |
| Q6_K | 2.4 GB | 3.8 GB | Fits | Get file → |
For 8 GB VRAM, use IQ1_S (0.68 GB, 2 GB memory at 8 k context) or IQ1_M (0.74 GB, 2.1 GB). For 12 GB VRAM, IQ2_S (0.96 GB, 2.3 GB) or Q2_K (1.12 GB, 2.5 GB) are good choices. For 24 GB VRAM, higher quality quants such as Q4_K_M (1.78 GB, 3.1 GB) or Q5_K_M (2.09 GB, 3.4 GB) can be used.
What GPU do you need for NinjaMouse 3B 40L Danube?
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: the best quant that fits is Q6_K (2.4 GB file, about 3.8 GB in use).
Which NinjaMouse 3B 40L Danube quant should you choose?
Q4_K_M is the usual starting point for NinjaMouse 3B 40L Danube: it balances file size and quality well. Here it is a 1.8 GB file.
If memory is tight, the smallest file is IQ1_S at 0.7 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
Source
Frequently asked questions about NinjaMouse 3B 40L Danube
How much VRAM does NinjaMouse 3B 40L Danube need?
About 3.1 GB at Q4_K_M with an 8K context window. That is the 1.8 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 NinjaMouse 3B 40L Danube on an 8 GB GPU?
Yes, with Q6_K: the 2.4 GB file needs about 3.8 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of NinjaMouse 3B 40L Danube GGUF?
It depends on the quant: IQ1_S is 0.7 GB, IQ1_M is 0.7 GB, IQ2_XXS is 0.8 GB, IQ2_XS is 0.9 GB, IQ2_S is 1.0 GB, IQ2_M is 1.0 GB and 18 more.
What context length does NinjaMouse 3B 40L Danube support?
The GGUF header lists a trained context length of 16,384 tokens (16K). The KV cache grows with the context you use: about 0.8 GB at 8K and 1.7 GB at 16K.
Where can I download NinjaMouse 3B 40L Danube GGUF?
The files are on Hugging Face in the mradermacher/NinjaMouse-3B-40L-danube-i1-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 NinjaMouse 3B 40L Danube 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.
Does this model support image generation?
No, the model has no vision or image generation capabilities.
What context length does it support?
It supports a maximum context length of 16384 tokens.