Back to news
launchLG2026-11-09

LG AI Research announced on November 9 that it is opening the K-EXAONE 2.5 model API for free to developers, enterprises and research institutions

The model is additionally trained on top of its August release, has agent capabilities comparable to leading open models, and performs on AutomationBench, SWE-Bench Pro, BrowseComp and Claw-Eval on par with DeepSeek-V4.1-Flash and GLM-5.3

On November 9, LG AI Research announced in Korea that it is opening the API for its K-EXAONE 2.5 model to the public for free. Any user, including developers, enterprises and research institutions, can apply the model to service development and research without paying.

The self developed technical verification model was additionally trained on top of its original release in August and has agent capabilities comparable to leading open artificial intelligence models, performing on AutomationBench, SWE-Bench Pro, BrowseComp and Claw-Eval on par with DeepSeek-V4.1-Flash and GLM-5.3.

Opening an API for free usually reflects a vendor trading real usage volume for ecosystem feedback and developer mindshare rather than revenue from the calls themselves. For research institutions and smaller developers, this kind of free allowance lowers the barrier to experimenting with new agent architectures and toolchains.

K-EXAONE 2.5LG免费接口智能体开源模型