South Korea
Organisations et modèles d’IA de South Korea
Organisations actives
3
Modèles suivis
35
Modèle le plus récent
sept. 2026
Dernières sorties de South Korea
septembre 2026
Upstage Solar Decide answers typed Noul, Choice, and Score questions against structured state.
août 2026
Solar Pro 4 is Upstage's agentic language model for multi-step tool use, terminal tasks, long-document reasoning, and multilingual work.
juillet 2026
K-EXAONE 2.0 is LG AI Research's 750B-parameter mixture-of-experts multilingual reasoning and agentic language model with 37B active parameters and a 262,144-token context window.
Solar Open 2 250B is Upstage's open-weight 250B-parameter hybrid-attention mixture-of-experts model with 15B active parameters and a 1M-token context window.
mars 2026
Solar Pro 3 is Upstage's 102B-parameter mixture-of-experts model, activating 12B parameters per token for multilingual reasoning and agentic workflows.
janvier 2026
Solar Pro 3 (2026-01-26) is the first dated release of Upstage's 102B-parameter, 12B-active mixture-of-experts model.
décembre 2025
Solar Open 100B is Upstage's open-weight 102.6B-parameter mixture-of-experts model with 12B active parameters per token.
K EXAONE is a language model from LG for chat, writing, analysis, and general-purpose automation. It accepts text inputs and produces text outputs.
HyperClova X Seed Omni 8B is a speech-to-text model from NAVER for transcription and spoken-language understanding workflows. It is suited to captioning, meeting notes, transcription, and speech analytics pipelines.
Organisations de South Korea
Modèles de South Korea
Parcourez le catalogue complet, classé par organisation.
35 modèles
LG
11 modèlesK-EXAONE 2.0 is LG AI Research's 750B-parameter mixture-of-experts multilingual reasoning and agentic language model with 37B active parameters and a 262,144-token context window.
K EXAONE is a language model from LG for chat, writing, analysis, and general-purpose automation. It accepts text inputs and produces text outputs.
EXAONE 4.0 1.2B is a language model from LG for chat, writing, analysis, and general-purpose automation. It accepts text inputs and produces text outputs.
EXAONE 4.0 32B is a language model from LG for chat, writing, analysis, and general-purpose automation. It accepts text inputs and produces text outputs.
EXAONE Deep 2.4B is a language model from LG for chat, writing, analysis, and general-purpose automation. It accepts text inputs and produces text outputs.
EXAONE Deep 32B is a language model from LG for chat, writing, analysis, and general-purpose automation. It accepts text inputs and produces text outputs.
NAVER
3 modèlesHyperClova X Seed Omni 8B is a speech-to-text model from NAVER for transcription and spoken-language understanding workflows. It is suited to captioning, meeting notes, transcription, and speech analytics pipelines.
HyperClova X Seed Think 32B is a reasoning model from NAVER for multi-step analysis, math, and tool-driven tasks. It accepts text inputs and produces text outputs.
HyperClova X Seed Think 14B is a reasoning model from NAVER for multi-step analysis, math, and tool-driven tasks. It accepts text inputs and produces text outputs.
Upstage
21 modèlesSolar Mini 4 is Upstage's compact, cost-efficient agentic language model, with 35B total parameters and 3B active parameters. It supports long-context chat, reasoning, structured outputs, and tool calling in Korean, English, and Japanese.
Upstage Solar Decide answers typed Noul, Choice, and Score questions against structured state.
Solar Pro 4 is Upstage's agentic language model for multi-step tool use, terminal tasks, long-document reasoning, and multilingual work.
Solar Open 2 250B is Upstage's open-weight 250B-parameter hybrid-attention mixture-of-experts model with 15B active parameters and a 1M-token context window.
Solar Pro 3 is Upstage's 102B-parameter mixture-of-experts model, activating 12B parameters per token for multilingual reasoning and agentic workflows.
Solar Pro 3 (2026-01-26) is the first dated release of Upstage's 102B-parameter, 12B-active mixture-of-experts model.