---
license: apache-2.0
library_name: llama.cpp
pipeline_tag: text-generation
base_model:
- Kwaipilot/KAT-Coder-V2.5-Dev
base_model_relation: quantized
quantized_by: RemySkye
tags:
- gguf
- llama.cpp
- quantized
- text-generation
- code
- coding
- agentic-coding
- mixture-of-experts
- moe
- qwen3.5
- qwen3.6
---

# KAT-Coder-V2.5-Dev-GGUF

GGUF files for [Kwaipilot/KAT-Coder-V2.5-Dev](https://huggingface.co/Kwaipilot/KAT-Coder-V2.5-Dev), including a BF16 reference GGUF and multiple `llama.cpp` quantizations.

## What is included

| File | Format | Size |
|---|---:|---:|
| `KAT-Coder-V2.5-Dev-BF16.gguf` | `BF16` | 64.61 GiB |
| `KAT-Coder-V2.5-Dev-Q8_0.gguf` | `Q8_0` | 34.37 GiB |
| `KAT-Coder-V2.5-Dev-Q6_K.gguf` | `Q6_K` | 26.56 GiB |
| `KAT-Coder-V2.5-Dev-Q5_K_M.gguf` | `Q5_K_M` | 23.03 GiB |
| `KAT-Coder-V2.5-Dev-Q5_1.gguf` | `Q5_1` | 24.32 GiB |
| `KAT-Coder-V2.5-Dev-Q5_0.gguf` | `Q5_0` | 22.33 GiB |
| `KAT-Coder-V2.5-Dev-Q5_K_S.gguf` | `Q5_K_S` | 22.33 GiB |
| `KAT-Coder-V2.5-Dev-Q4_K_M.gguf` | `Q4_K_M` | 19.71 GiB |
| `KAT-Coder-V2.5-Dev-Q4_1.gguf` | `Q4_1` | 20.35 GiB |
| `KAT-Coder-V2.5-Dev-Q4_0.gguf` | `Q4_0` | 18.36 GiB |
| `KAT-Coder-V2.5-Dev-Q4_K_S.gguf` | `Q4_K_S` | 18.52 GiB |
| `KAT-Coder-V2.5-Dev-IQ4_XS.gguf` | `IQ4_XS` | 17.64 GiB |
| `KAT-Coder-V2.5-Dev-Q3_K_L.gguf` | `Q3_K_L` | 16.87 GiB |
| `KAT-Coder-V2.5-Dev-Q3_K_M.gguf` | `Q3_K_M` | 15.61 GiB |
| `KAT-Coder-V2.5-Dev-Q3_K_S.gguf` | `Q3_K_S` | 14.14 GiB |
| `KAT-Coder-V2.5-Dev-Q2_K.gguf` | `Q2_K` | 12.05 GiB |

All model behavior, intended use, limitations, and licensing originate from [Kwaipilot/KAT-Coder-V2.5-Dev](https://huggingface.co/Kwaipilot/KAT-Coder-V2.5-Dev).