Gemini Embedding: Generalizable Embeddings from Gemini
This provides a unified embedding solution for multilingual and multimodal text tasks, though it is incremental as it builds on existing large language model capabilities.
The paper tackles the problem of creating generalizable text embeddings across multiple languages and modalities by introducing Gemini Embedding, which leverages Google's Gemini model to achieve state-of-the-art performance on the Massive Multilingual Text Embedding Benchmark, outperforming prior models.
In this report, we introduce Gemini Embedding, a state-of-the-art embedding model leveraging the power of Gemini, Google's most capable large language model. Capitalizing on Gemini's inherent multilingual and code understanding capabilities, Gemini Embedding produces highly generalizable embeddings for text spanning numerous languages and textual modalities. The representations generated by Gemini Embedding can be precomputed and applied to a variety of downstream tasks including classification, similarity, clustering, ranking, and retrieval. Evaluated on the Massive Multilingual Text Embedding Benchmark (MMTEB), which includes over one hundred tasks across 250+ languages, Gemini Embedding substantially outperforms prior state-of-the-art models, demonstrating considerable improvements in embedding quality. Achieving state-of-the-art performance across MMTEB's multilingual, English, and code benchmarks, our unified model demonstrates strong capabilities across a broad selection of tasks and surpasses specialized domain-specific models.