Text

Kanana

Compact Korean-English models from Kakao. Kanana 2 30B-A3B is fast thanks to MoE; small 1–3B versions suit a regular PC.

Developer
Kakao, South Korea
First release
Feb 2025
Latest release
Jul 2026
Sizes
1.3B – 30B-A3B
License
Commercial use with conditionsKanana 1.5: Apache 2.0; Kanana 2: Kakao's own license (companies with more than 10 million monthly users need a separate license); Kanana Nano: CC-BY-NC
Russian
Not supported
Ready-made builds
GGUF
Running
On your own serverAlso runs without a GPU
Industries
Customer support
Deploy on your server Hugging Face GitHub

What it does

  • Support chatbot
  • Customer request classification
  • Lightweight assistant on your own PC

Where it is used

Customer supportMessengers and servicesWork with Korean partners

Hardware requirements

LaptopLaptop or regular PC, up to 8 GB of VRAM — smaller versions
fits
1 GPUOne GPU with 16–80 GB — mid-size versions
fits
ClusterServer with several GPUs — flagship versions
no versions

Versions

  1. Kanana 2 1.3B / 3B
  2. Kanana 2 30B-A3B
  3. Kanana 1.5 15.7B-A3B
  4. Kanana 1.5 (2.1B, 8B)
  5. Kanana Nano 2.1B

How to run it

On your own serverWeights are downloaded from Hugging Face and served with vLLM (load and API) or llama.cpp (modest hardware). The model then runs in a closed network with no per-request fees.Hugging Face
How much hardware you needCalculate the VRAM for the model size, context length and number of concurrent requests.Open the hardware calculator

I can set this up end to end: pick the model size, deploy it on your server and connect it to your systems. Quantization compresses a model so it takes less video memory and runs on more modest hardware. Answers change slightly, so quality is checked on your own examples.

Frequently asked questions

Can Kanana be used in a commercial project?

With conditions. License: Kanana 1.5: Apache 2.0; Kanana 2: Kakao's own license (companies with more than 10 million monthly users need a separate license); Kanana Nano: CC-BY-NC. Restrictions vary — region, company revenue, attribution requirements. Have a lawyer check the terms before a commercial launch.

What hardware does Kanana need?

At minimum: Laptop or regular PC, up to 8 GB of VRAM — smaller versions. Some versions also run on an ordinary CPU, without a GPU. You can calculate the exact VRAM for your model size and context in the hardware calculator.

Does Kanana support Russian?

No. The model card lists its languages and Russian is not among them.

Where can I download Kanana and what does it cost?

The Kanana weights are open and free to download. You only pay for the hardware it runs on and for the setup. Source links are at the bottom of this page.

How I deploy it for clients

  1. SelectionI pick the model size for your task and hardware and test it on your examples.
  2. DeploymentI deploy it on your server or in a closed network and provide an API.
  3. Fine-tuningI fine-tune it on your data (LoRA) or connect a knowledge base — whichever is cheaper for the task.
  4. IntegrationI connect it to your CRM, ERP, bot, website or team chat and set up monitoring.

Similar models

TextEXAONELG AI Research · South KoreaCommercial use with conditions

Korean-English models from LG. Most of the line is non-commercial, but the flagship K-EXAONE 2.0 with 750 billion parameters is released under Apache 2.0.

DetailsTextHyperCLOVA X SEEDNaver · South KoreaCommercial use with conditions

Open smaller models from Korea's Naver: from 0.5B to 32B, including reasoning Think versions and multimodal versions that understand images.

DetailsTextSolarUpstage · South KoreaCommercial use with conditions

Models from Korea's Upstage. Solar Open 2 is built for office document work: 250 billion parameters, 15 billion active; languages are English, Korean and Japanese.

Details

Source: huggingface.co/kakaocorp/kanana-2-3b-instruct. Data checked against the model card on 22 Sep 2026. Have a lawyer review the license before commercial launch.