KA

KAT Coder V2.5 Dev MTP

KatGGUF by gbuzhfUpdated 2026-09-12
Download on Hugging Face
Parameters35.5B
Context256K
Layers41
Architectureqwen35moe
Smallest quant14.4 GB
Quick answer

KAT Coder V2.5 Dev MTP needs about 24.0 GB of memory at Q4_K_XL (a 22.9 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 9 quants, from 14.4 GB to 30.0 GB (Q6_K). Longer context needs more: at 32K tokens the KV cache alone adds about 2.8 GB.

KAT Coder V2.5 Dev MTP – a 37B Qwen3.5‑MOE coding model in GGUF format

KAT Coder V2.5 Dev MTP is a 37‑billion‑parameter Qwen3.5‑MOE model optimized for coding tasks. It is released under the Apache‑2.0 license and ships only text‑only weights; no vision or multimodal components are included. The model is available as a set of GGUF files with various quantizations.

Good at

  • High‑quality code generation
  • Large context handling up to 262,144 tokens
  • Fine‑tuned for agentic coding scenarios

Limits

  • Requires large VRAM for full precision (≈15–31 GB at 8 k context)
  • No vision or multimodal support
  • Limited to text‑only usage

Best for

  • Developing coding assistants
  • Running code‑generation benchmarks
  • Fine‑tuning for specialized coding tasks

Who it’s for. Engineers and researchers who need a large‑scale, text‑only coding model with a 262,144‑token context window. The model is suitable for developers building agentic coding tools or running extensive code‑generation benchmarks.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—14.4 GB 15.6 GB
Too big Get file →
—17.3 GB 18.5 GB
Too big Get file →
—17.4 GB 18.6 GB
Too big Get file →
IQ4_XS18.2 GB 19.4 GB
Too big Get file →
Q4_K_XL22.9 GB 24.0 GB
Too big Get file →
—23.7 GB 24.9 GB
Too big Get file →
Q5_K_S25.5 GB 26.7 GB
Too big Get file →
—26.2 GB 27.4 GB
Too big Get file →
Q6_K30.0 GB 31.2 GB
Too big Get file →

For 8 GB VRAM, use the Q4_K_XL quantization (22.85 GB file, ~24 GB memory at 8 k context). For 12 GB VRAM, Q5_K_S (25.54 GB file, ~26.7 GB memory) is a good balance. For 16 GB or more, the 30.01 GB Q6_K file provides the highest quality with acceptable memory usage. Memory estimates are approximate.

What GPU do you need for KAT Coder V2.5 Dev MTP?

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: the best quant that fits is the main file (14.4 GB file, about 15.6 GB in use).
  • 24 GB VRAM: the best quant that fits is Q4_K_XL (22.9 GB file, about 24.0 GB in use).
  • 32 GB VRAM: the best quant that fits is Q6_K (30.0 GB file, about 31.2 GB in use).

Which KAT Coder V2.5 Dev MTP quant should you choose?

Q4_K_XL (22.9 GB) is the middle option in this repository. The usual sweet spot between file size and quality, and the quant most people start with.

If memory is tight, the smallest file is the first one in the table at 14.4 GB, with the quality trade-offs that come with fewer bits.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureqwen35moe
Parameters35.5B (approximate, mixture of experts)
Layers41
Trained context262,144 tokens
KV cache (fp16)82 KB per token

Source

Quantized bygbuzhf
FamilyKat
Licenseapache-2.0
Downloads (HF)1,318
Last updated2026-09-12

Frequently asked questions about KAT Coder V2.5 Dev MTP

How much VRAM does KAT Coder V2.5 Dev MTP need?

About 24.0 GB at Q4_K_XL with an 8K context window. That is the 22.9 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 KAT Coder V2.5 Dev MTP on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 15.6 GB, so you would need a larger GPU or CPU offloading.

What is the file size of KAT Coder V2.5 Dev MTP GGUF?

It depends on the quant: file is 14.4 GB, file is 17.3 GB, file is 17.4 GB, IQ4_XS is 18.2 GB, Q4_K_XL is 22.9 GB, file is 23.7 GB and 3 more.

What context length does KAT Coder V2.5 Dev MTP 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 2.8 GB at 32K.

Where can I download KAT Coder V2.5 Dev MTP GGUF?

The files are on Hugging Face in the gbuzhf/KAT-Coder-V2.5-Dev-MTP-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 MTP 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 MTP on 8 GB VRAM?

Only with a low‑bit quant like Q4_K_XL; the file is 22.85 GB and requires about 24 GB of VRAM at 8 k context.

Does this model support image input?

No, the released weights are text‑only; vision components are not included.

What license applies to this model?

The model is released under the Apache‑2.0 license.

Other GGUFs of KAT-Coder-V2.5-Dev

Similar models

Browse all →

More from gbuzhf

All by gbuzhf →

More Kat models

All Kat →

Other large models

All large models →