KI

Kimi-K2.7-Code

KimiGGUF by unslothUpdated 2026-06-19
Download on Hugging Face
Parameters1030B
Context256K
Layers61
Architecturedeepseek2
Smallest quant303.9 GB
Quick answer

Kimi-K2.7-Code needs about 420.4 GB of memory at IQ3_S (a 418.8 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 10 quants, from 303.9 GB (IQ1_M) to 594.5 GB (Q8_K_XL). Longer context needs more: at 32K tokens the KV cache alone adds about 4.3 GB.

Kimi-K2.7-Code: a large coding‑focused agentic model

Kimi-K2.7-Code is a 950 billion‑parameter model built on the deepseek2 architecture. It excels at long‑horizon coding tasks and supports vision. The model was trained with a 262,144‑token context length.

Good at

  • Long‑horizon coding
  • Agentic task completion
  • Vision‑enabled code analysis

Limits

  • Requires large VRAM for full precision
  • No native vision support in llama.cpp
  • License is other, not open source

Best for

  • Complex software engineering workflows
  • End‑to‑end code generation
  • Vision‑augmented coding tasks

Who it’s for. Developers and research teams needing a powerful coding agent with vision capabilities. Suitable for teams that can allocate sufficient GPU memory and are comfortable with a non‑open‑source license.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ1_M303.9 GB (8 parts) 305.5 GB
Too big Get file →
IQ2_XXS317.8 GB (8 parts) 319.4 GB
Too big Get file →
IQ2_M318.0 GB (8 parts) 319.6 GB
Too big Get file →
Q2_K_XL339.5 GB (8 parts) 341.1 GB
Too big Get file →
IQ3_S418.8 GB (10 parts) 420.4 GB
Too big Get file →
Q3_K_M463.6 GB (11 parts) 465.2 GB
Too big Get file →
Q3_K_XL463.9 GB (11 parts) 465.5 GB
Too big Get file →
IQ4_XS495.1 GB (12 parts) 496.7 GB
Too big Get file →
Q4_K_XL583.7 GB (14 parts) 585.3 GB
Too big Get file →
Q8_K_XL594.5 GB (14 parts) 596.1 GB
Too big Get file →

For 8 GB VRAM, use IQ1_M or IQ2_XXS (≈305 GB file size, 305 GB memory at 8k context). For 16 GB VRAM, Q2_M or Q3_K_M are suitable. For 32 GB VRAM, Q4_K_XL or Q8_K_XL provide higher precision while staying within memory limits. Estimates are approximate.

What GPU do you need for Kimi-K2.7-Code?

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: no quant fits at an 8K context.
  • 32 GB VRAM: no quant fits at an 8K context.
  • 48 GB VRAM: no quant fits at an 8K context.

Which Kimi-K2.7-Code quant should you choose?

IQ3_S (418.8 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.

If memory is tight, the smallest file is IQ1_M at 303.9 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

Architecturedeepseek2
Parameters1030B (approximate, mixture of experts)
Layers61
Trained context262,144 tokens
KV cache (fp16)130 KB per token

Source

Quantized byunsloth
FamilyKimi
Licenseother
Downloads (HF)145,948
Last updated2026-06-19

Frequently asked questions about Kimi-K2.7-Code

How much VRAM does Kimi-K2.7-Code need?

About 420.4 GB at IQ3_S with an 8K context window. That is the 418.8 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 Kimi-K2.7-Code on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 305.5 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Kimi-K2.7-Code GGUF?

It depends on the quant: IQ1_M is 303.9 GB, IQ2_XXS is 317.8 GB, IQ2_M is 318.0 GB, Q2_K_XL is 339.5 GB, IQ3_S is 418.8 GB, Q3_K_M is 463.6 GB and 4 more.

What context length does Kimi-K2.7-Code support?

The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 1.1 GB at 8K and 4.3 GB at 32K.

Where can I download Kimi-K2.7-Code GGUF?

The files are on Hugging Face in the unsloth/Kimi-K2.7-Code-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 Kimi-K2.7-Code use?

Hugging Face lists the license as other. 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 Kimi-K2.7-Code on 8 GB VRAM?

Only the lowest‑precision quants (IQ1_M, IQ2_XXS) can fit, but they still require around 305 GB of VRAM at 8k context, so 8 GB is insufficient.

Does this model support vision in llama.cpp?

The model has vision abilities, but llama.cpp does not natively support vision for this model.

What is the context length of this model?

The model was trained with a 262,144‑token context length.

Other GGUFs of Kimi-K2.7-Code

Similar models

Browse all →

More from unsloth

All by unsloth →

More Kimi models

All Kimi →

Other large models

All large models →