Mark-Engineer-V3-1.5B needs about 1.5 GB of memory at IQ4_NL (a 0.9 GB file) with an 8K context window, so it fits on a 4 GB GPU. Longer context needs more: at 32K tokens the KV cache alone adds about 0.7 GB.
Mark-Engineer-V3-1.5B – a 1.5B parameter systems engineering LLM
Mark-Engineer-V3-1.5B is a 1,520,496,960 parameter model with 405.73M active parameters per token. It uses 32 routed experts, 2 shared experts, and domain soft‑pinning for code generation, compilers, and kernel internals. The model supports a 32,768 token context.
Good at
- Code generation
- Compiler assistance
- Kernel internals
Limits
- Context length 32,768 tokens
- Estimated 1.5 GB VRAM at 8 k context
- AGPL‑3.0 license
Best for
- Developers writing low‑level code
- Compiler researchers
- Kernel developers
Who it’s for. Engineers and researchers needing a specialized LLM for systems‑level code tasks. The model is tailored for environments where code generation and low‑level language understanding are critical.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| IQ4_NL | 0.9 GB | 1.5 GB | Fits | Get file → |
The only available quantization is IQ4_NL, 0.86 GB in size. It requires approximately 1.5 GB of VRAM at an 8 k context. This fits comfortably on 8 GB GPUs and is suitable for most code‑generation workloads.
What GPU do you need for Mark-Engineer-V3-1.5B?
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 IQ4_NL (0.9 GB file, about 1.5 GB in use).
Which Mark-Engineer-V3-1.5B quant should you choose?
IQ4_NL (0.9 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
Source
Frequently asked questions about Mark-Engineer-V3-1.5B
How much VRAM does Mark-Engineer-V3-1.5B need?
About 1.5 GB at IQ4_NL with an 8K context window. That is the 0.9 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 Mark-Engineer-V3-1.5B on an 8 GB GPU?
Yes, with IQ4_NL: the 0.9 GB file needs about 1.5 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Mark-Engineer-V3-1.5B GGUF?
It depends on the quant: IQ4_NL is 0.9 GB.
What context length does Mark-Engineer-V3-1.5B support?
The GGUF header lists a trained context length of 32,768 tokens (32K). The KV cache grows with the context you use: about 0.2 GB at 8K and 0.7 GB at 32K.
Where can I download Mark-Engineer-V3-1.5B GGUF?
The files are on Hugging Face in the miceworks/Mark-Engineer-V3-1.5B repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does Mark-Engineer-V3-1.5B use?
Hugging Face lists the license as agpl-3.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
How many experts does the model use?
It uses 32 routed experts and 2 shared experts.
What license does the model use?
The model is released under the AGPL‑3.0 license.