Multilingual Semantic Retrieval for Apple Music Search
Apple ML Research
Apple Music developed a multilingual semantic retrieval system using a 305M-parameter Siamese bi-encoder model fine-tuned from GTE-multilingual-base to improve search across 150+ storefronts and dozens of languages. In an online A/B test, the system achieved a 2.28% relative conversion-rate lift overall, an 86% reduction in the no-result rate, and a 7.93% relative CR lift for tail queries. The hybrid retrieval architecture integrates dense nearest-neighbor results with token-based indexing, enabling improved recall on misspelled, transliterated, and cross-lingual queries without retraining downstream rankers.
Why it matters
Apple Music serves listeners across 150+ storefronts in dozens of languages, with a catalog that grows by hundreds of thousands of new tracks daily. At this scale, search recall on misspelled, transliterated, and cross-lingual queries becomes a dominant driver of session quality, particularly for tail queries that account for the majority of unique queries. We present a multilingual semantic retrieval system built on a 305M-parameter Siamese bi-encoder fine-tuned from GTE-multilingual-base with curriculum-scheduled multi-objective training. The model is integrated into the search stack via a…
