We present the MultiplEYE Text Corpus, a large-scale, document-level, multi-parallel resource designed to advance cross-linguistic research on reading and language processing. The corpus provides paragraph-level alignment for texts in 39 languages spanning seven language families and seven scripts. Unlike many existing multilingual corpora, a substantial number of documents were originally written in languages other than English, reducing English-centric bias and supporting more typologically diverse investigations. The texts are carefully selected to balance linguistic richness with experimental feasibility, particularly for eye-tracking-while-reading studies. Developed within a multi-lab initiative, the MultiplEYE Text Corpus follows unified translation, alignment, and experimental design guidelines to ensure cross-linguistic comparability. Its inclusion of texts varying in type and difficulty enables research on discourse- level processing, genre effects, and individual differences across a wide range of languages. The text corpus and accompanying metadata provide a robust foundation for multilingual psycholinguistic and computational modeling research. Data and materials are publicly available at https://doi.org/10.23668/psycharchives.22294.

Long-video understanding on edge devices must reason over hours of content under tight compute and bandwidth budgets. Subsampling visual tokens loses temporal structure, while text-only video memories lose fine-grained visual attributes. We observe a visual-textual duality: language memories carry long-range temporal structure better than dense frames, while pixels remain decisive for attribute-level perception. Building on this insight, we propose Caption-once, Frames-onDemand (CFD), a budget-aware edge-cloud agentic framework. The edge runs a single offline captioning pass that builds a dual-track narrative index, an event-level story skeleton plus a clip-level micro-log, cached and reused across queries without re-captioning. At query time, a cloud-side MLLM reasons over the index in a story-first loop centered on a lightweight Visual-Need Router: a per-query gating module that triggers bounded keyframe retrieval only for perceptual questions (appearance, on-screen text, attribute disambiguation) and keeps temporal-structural questions in language space. The router turns visual access into a first-class, query-conditioned cost, capping per-query frame consumption regardless of video length. Experiments on long-video benchmarks demonstrate strong accuracy-efficiency trade-offs while substantially reducing online visual processing.