Open web-scale pre-training corpora remain concentrated in English, limiting multilingual LLM development. We introduce MultiSynt/MT, an open synthetic parallel corpus with approximately 4.8 trillion target-language tokens across 36 European languages, produced by translating 100 billion high-quality Nemotron-CC tokens with Tower+ and OPUS-MT/HPLT-MT systems. For many medium- and lower-resource European languages, this is the largest openly available pre-training resource. On a broad multilingual benchmark suite, reference LLMs trained on MultiSynt/MT reach the final score of HPLT 2.0, a native-data baseline, using roughly 72% fewer pre-training tokens, and outperform it by approximately 15% relative at a matched 100B-token training budget. Our analyses also identify evaluation blind spots: standard multiple-choice benchmarks miss translation-quality differences that a fluency-sensitive LLM-as-judge evaluation cleanly recovers on the trained LLMs (with no fluency deficit in MultiSynt itself), and Norwegian idiomatic and culturally grounded tasks remain better served by native data. We release the corpus, including row-aligned translations from multiple systems, to support controlled research on multilingual pre-training data and evaluation.

Considering the rapid improvement of large generative language models, it is important to measure their ability to encode clinical domain knowledge in order to help determine their potential utility in a clinical setting. To this end we present MedQA-SWE – a novel multiple choice, clinical question & answering (Q&A) dataset in Swedish consisting of 3,180 questions. The dataset was created from a series of exams aimed at evaluating doctors’ clinical understanding and decision making and is the first open-source clinical Q&A dataset in Swedish. The exams – originally in PDF format – were parsed and each question manually checked and curated in order to limit errors in the dataset. We provide dataset statistics along with benchmark accuracy scores of seven large generative language models on a representative sample of questions in a zero-shot setting, with some models showing impressive performance given the difficulty of the exam the dataset is based on.