EXAONE 4.5 Technical Report

TL;DR

EXAONE 4.5 is an open-weight vision language model by LG AI Research, integrating visual and textual modalities, with enhanced document understanding, long context handling, and competitive benchmark performance.

Contribution

It introduces a multimodal architecture with a visual encoder into EXAONE 4.0, emphasizing document-centric training for improved industrial and language tasks.

Findings

Achieves substantial gains in document understanding and Korean reasoning.

Extends context length up to 256K tokens for long-context reasoning.

Outperforms state-of-the-art models of similar scale in key benchmarks.

Abstract

This technical report introduces EXAONE 4.5, the first open-weight vision language model released by LG AI Research. EXAONE 4.5 is architected by integrating a dedicated visual encoder into the existing EXAONE 4.0 framework, enabling native multimodal pretraining over both visual and textual modalities. The model is trained on large-scale data with careful curation, particularly emphasizing document-centric corpora that align with LG's strategic application domains. This targeted data design enables substantial performance gains in document understanding and related tasks, while also delivering broad improvements across general language capabilities. EXAONE 4.5 extends context length up to 256K tokens, facilitating long-context reasoning and enterprise-scale use cases. Comparative evaluations demonstrate that EXAONE 4.5 achieves competitive performance in general benchmarks while outperforming state-of-the-art models of similar scale in document understanding and Korean contextual reasoning. As part of LG's ongoing effort toward practical industrial deployment, EXAONE 4.5 is designed to be continuously extended with additional domains and application scenarios to advance AI for a better life.