KA

Kanana 2 30b a3b instruct 2601

KananaGGUF by deulUpdated 2026-07-04
Download on Hugging Face
Parameters30.7B
Context32K
Layers48
Architecturedeepseek2
Smallest quant18.6 GB
Quick answer

Kanana 2 30b a3b instruct 2601 needs about 20.0 GB of memory at Q4_K_M (a 18.6 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 4 quants, from 18.6 GB (Q4_K_M) to 32.6 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 3.4 GB.

Kanana 2 30b a3b instruct 2601 – 30.7B deepseek2 model with 32K context

Kanana 2 30b a3b instruct 2601 is a 30.7B parameter deepseek2 model with MoE and MLA architecture. It supports 32,768 token context and is tuned for instruction following and tool calling.

Good at

  • Tool calling
  • Reasoning
  • Long‑context instruction following

Limits

  • Requires 20+ GB VRAM for Q4_K_M
  • No vision support
  • License is Kanana License

Best for

  • Agentic AI agents
  • Complex instruction workflows
  • Long document analysis

Who it’s for. Developers building agentic systems that need robust reasoning and tool usage. Suitable for applications requiring extensive context handling and multilingual support.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_K_M18.6 GB 20.0 GB
Too big Get file →
Q5_K_M21.8 GB 23.2 GB
Too big Get file →
Q6_K25.2 GB 26.6 GB
Too big Get file →
Q8_032.6 GB 34.0 GB
Too big Get file →

For 20 GB VRAM use Q4_K_M (file size 18.6 GB, estimated 20 GB memory). For 23.2 GB VRAM use Q5_K_M (21.8 GB). For 26.6 GB VRAM use Q6_K (25.23 GB). For 34 GB VRAM use Q8_0 (32.63 GB). Estimates are approximate.

What GPU do you need for Kanana 2 30b a3b instruct 2601?

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 Q5_K_M (21.8 GB file, about 23.2 GB in use).
  • 32 GB VRAM: the best quant that fits is Q6_K (25.2 GB file, about 26.6 GB in use).
  • 48 GB VRAM: the best quant that fits is Q8_0 (32.6 GB file, about 34.0 GB in use).

Which Kanana 2 30b a3b instruct 2601 quant should you choose?

Q4_K_M is the usual starting point for Kanana 2 30b a3b instruct 2601: it balances file size and quality well. Here it is a 18.6 GB file.

The Q8_0 file is 32.6 GB, about 1.8 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturedeepseek2
Parameters30.7B (approximate, mixture of experts)
Layers48
Trained context32,768 tokens
KV cache (fp16)102 KB per token

Source

Quantized bydeul
FamilyKanana
Licenseother
Downloads (HF)497
Last updated2026-07-04

Frequently asked questions about Kanana 2 30b a3b instruct 2601

How much VRAM does Kanana 2 30b a3b instruct 2601 need?

About 20.0 GB at Q4_K_M with an 8K context window. That is the 18.6 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 Kanana 2 30b a3b instruct 2601 on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 20.0 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Kanana 2 30b a3b instruct 2601 GGUF?

It depends on the quant: Q4_K_M is 18.6 GB, Q5_K_M is 21.8 GB, Q6_K is 25.2 GB, Q8_0 is 32.6 GB.

What context length does Kanana 2 30b a3b instruct 2601 support?

The GGUF header lists a trained context length of 32,768 tokens (32K). The KV cache grows with the context you use: about 0.9 GB at 8K and 3.4 GB at 32K.

Where can I download Kanana 2 30b a3b instruct 2601 GGUF?

The files are on Hugging Face in the deul/Kanana-2-30b-a3b-instruct-2601-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 Kanana 2 30b a3b instruct 2601 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 this model on 8 GB VRAM?

No, the smallest quantization requires at least 20 GB VRAM.

Does this model support vision?

No, it does not have vision capabilities.

How many languages does it support?

It supports Korean, English, Japanese, Chinese, Thai, and Vietnamese.

Other GGUFs of kanana-2-30b-a3b-instruct-2601

Similar models

Browse all →

More Kanana models

All Kanana →

Other medium models

All medium models →