AI NewsWaveKakao Releases Two Embedding Models for Personalized AI Agents as Open Source
Key points
- Kakao releases two Kanana-2 embedding models
- 3B and 1.3B models, open source on Hugging Face
- Foundation technology for Korean search and personalized agents
Kakao announced on the 29th that it has released two embedding models, "Kanana-2 Embedding," as open source on the AI development platform Hugging Face, as foundation technology for personalized AI agents. The models are "Kanana-2-3B-Embedding," with 3 billion parameters, and "Kanana-2-1.3B-Embedding," with 1.3 billion parameters.
The uses are search and agents. Digital Today reported that the models can be used for personalized AI agents and meaning-based search, and Chosun Biz introduced them as models with stronger Korean-language search that support information retrieval by personalized AI agents. Chosun Biz added that this follows Kakao's introduction of four small language models (SLMs) in July.
Glossary Embedding is a technology that turns the meaning of a sentence into numbers so a computer can calculate how close or far apart meanings are.