cd /news/artificial-intelligence/lg-research-releases-k-exaone-2-5-ap… · home › topics › artificial-intelligence › article
[ARTICLE · art-168170] src=lgresearch.ai ↗ pub= topic=artificial-intelligence verified=true sentiment=↑ positive

LG Research releases K-Exaone 2.5 API for free

LG AI Research released a free API for K-EXAONE 2.5, its proprietary foundation model, through the EXAONE API Platform at docs.exaone.ai, available to developers, startups, enterprises and research institutions via an API key. K-EXAONE 2.5 is a technical validation model given additional training on top of the version released last August, and in internal evaluations on publicly available datasets it scored 74.6 on AutomationBench-Openset, 59.2 on SWE-Bench Pro, 78.0 on BrowseComp and 77.3 on Claw-Eval, which LG AI Research describes as competitive with frontier-class open-weight models including DeepSeek-V4.1-Flash and GLM-5.3.

read2 min views1 publishedNov 12, 2026
LG Research releases K-Exaone 2.5 API for free
Image: source

LG AI Research is making the API for K-EXAONE 2.5, its proprietary AI foundation model, available for free. K-EXAONE 2.5 is a technical validation model that has undergone additional training based on the version released in last August. The K-EXAONE 2.5 API is available for free through the EXAONE API Platform(docs.exaone.ai). Developers, startups, enterprises, research institutions, and anyone else can use K-EXAONE 2.5 for service development and research at no cost simply by obtaining an API key.

K-EXAONE 2.5 Demonstrates Competitive Agentic AI Performance Against Frontier-Class Models

(Based on internal evaluations using publicly available datasets, K-EXAONE 2.5 demonstrated competitive Agentic AI performance against frontier-class models.)

K-EXAONE 2.5 demonstrates strong capabilities in Agentic AI. In internal benchmark evaluations conducted using publicly available datasets, the model delivered performance comparable to global frontier-class open-weight AI models. These capabilities were demonstrated across four benchmarks evaluating Agentic AI performance. Notably, K-EXAONE 2.5 achieved competitive performance compared to leading Chinese frontier-class models, including DeepSeek-V4.1-Flash and GLM-5.3.

●AutomationBench-Openset - 74.6: Evaluates the ability to automate tasks using SaaS tools.

●SWE-Bench Pro - 59.2: Evaluates the ability to perform real-world software development and modification tasks.

●BrowseComp - 78.0: Evaluates the ability to navigate the web and find relevant information.

●Claw-Eval - 77.3: Evaluates an AI agent's problem-solving capabilities in workplace environments.

Boost Software Development Productivity with K-EXAONE 2.5

From intelligent AI agents to advanced coding assistance, K-EXAONE 2.5 offers powerful capabilities in complex reasoning, coding, and workflow automation, helping developers turn new ideas into real-world applications. In particular, K-EXAONE 2.5 can help boost software development productivity.From code generation and debugging to refactoring and test creation, K-EXAONE 2.5 helps developers tackle complex coding challenges, streamline development workflows, and build applications more quickly and efficiently.

Explore a wide range of possibilities with K-EXAONE 2.5. Your ideas could be the beginning of new innovations. Start building with K-EXAONE 2.5 today. We can't wait to see what you create!

── more in #artificial-intelligence 4 stories · sorted by recency
── more on @lg ai research 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/lg-research-releases…] indexed:0 read:2min 2026-11-12 · —