KAT Coder V2.5 Dev 16GB MixedQuant needs about 16.2 GB of memory at IQ3_XXS (a 15.0 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 2 quants, from 15.0 GB (IQ3_XXS) to 19.5 GB (IQ4_XS). Longer context needs more: at 32K tokens the KV cache alone adds about 2.7 GB.
KAT Coder V2.5 Dev 16GB MixedQuant – a 34.7B Qwen3.5‑moe coding model
KAT Coder V2.5 Dev is a 34.7B parameter Qwen3.5‑moe model fine‑tuned for coding tasks. The mixed‑quantization build balances size and speed, with IQ3_XXS and IQ4_XS variants. It is text‑only and licensed under Apache‑2.0.
Good at
- Coding assistance
- Code generation
- SFT/RL‑trained agentic coding
Limits
- No vision or multimodal support
- Requires 16‑24GB VRAM for IQ4_XS
- Large context needs 22‑24GB VRAM
Best for
- Developers needing fast code generation
- Agents for coding benchmarks
- Local inference on 16‑24GB GPUs
Who it’s for. Engineers and researchers who want a large‑scale coding model that can run locally on a single GPU. The model is open‑weight and suitable for integration into custom inference pipelines.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| IQ3_XXS | 15.0 GB | 16.2 GB | Too big | Get file → | |
| IQ4_XS | 19.5 GB | 20.6 GB | Too big | Get file → |
For 16GB VRAM, use the IQ3_XXS quantization (file size 14.99 GB, estimated 16.2 GB memory at 8K context). If you have 22‑24GB VRAM, the IQ4_XS quantization (19.45 GB, 20.6 GB memory at 8K context) offers higher precision. Both are approximate estimates.
What GPU do you need for KAT Coder V2.5 Dev 16GB MixedQuant?
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: the best quant that fits is IQ4_XS (19.5 GB file, about 20.6 GB in use).
Which KAT Coder V2.5 Dev 16GB MixedQuant quant should you choose?
IQ3_XXS (15.0 GB) is the middle option in this repository. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about KAT Coder V2.5 Dev 16GB MixedQuant
How much VRAM does KAT Coder V2.5 Dev 16GB MixedQuant need?
About 16.2 GB at IQ3_XXS with an 8K context window. That is the 15.0 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 KAT Coder V2.5 Dev 16GB MixedQuant on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 16.2 GB, so you would need a larger GPU or CPU offloading.
What is the file size of KAT Coder V2.5 Dev 16GB MixedQuant GGUF?
It depends on the quant: IQ3_XXS is 15.0 GB, IQ4_XS is 19.5 GB.
What context length does KAT Coder V2.5 Dev 16GB MixedQuant support?
The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 0.7 GB at 8K and 2.7 GB at 32K.
Where can I download KAT Coder V2.5 Dev 16GB MixedQuant GGUF?
The files are on Hugging Face in the fooooooooooooooooooooooooooooooooooooooooo/KAT-Coder-V2.5-Dev-16GB-MixedQuant-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 KAT Coder V2.5 Dev 16GB MixedQuant 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 KAT Coder V2.5 Dev on a 16GB GPU?
Yes, the IQ3_XXS quantization is designed to fit on 16GB VRAM, with an estimated 16.2GB memory usage at 8K context.
Does the model support image or multimodal input?
No, the model is text‑only; vision components are not included.
What license applies to this model?
The model is released under the Apache‑2.0 license.