Elastic Announces Elastic Rerank Model to Power Up Semantic Search

New cross-encoder reranking model amplifies search experiences through semantic boosting and no required reindexing

SAN FRANCISCO--(BUSINESS WIRE)--Elastic (NYSE: ESTC), the Search AI Company, announced Elastic Rerank, a cross-encoder reranking model that offers powerful semantic search capabilities with no required reindexing, and high relevance, top performance, and efficiency for text search. Developers can now semantically boost keyword search with little to no change to how data is indexed and searched, providing flexibility and control over costs.

“Reranking models provide a semantic boost to any search experience,” said Steve Kearns, general manager, Search at Elastic. “Building a reranking model into the Elasticsearch Open Inference API makes Elastic Rerank effortless to load and use in search pipelines. It allows users to quickly apply the accuracy benefits of semantic ranking to their Elasticsearch data just by adding a few parameters to existing queries.”

Built on the DeBERTa v3 architecture, the Elastic Rerank model outperforms other significantly larger reranking models. Testing indicates a 40% uplift on a broad range of retrieval tasks and up to 90% on question-answering data sets.

Support for Elastic Rerank via the Inference API is available on Elasticsearch Serverless and in Elasticsearch 8.17. Read the Elastic blog for more details.

About Elastic

Elastic (NYSE: ESTC), the Search AI Company, enables everyone to find the answers they need in real-time using all their data, at scale. Elastic’s solutions for search, observability, and security are built on the Elastic Search AI Platform, the development platform used by thousands of companies, including more than 50% of the Fortune 500. Learn more at elastic.co.

Elastic and associated marks are trademarks or registered trademarks of Elastic N.V. and its subsidiaries. All other company and product names may be trademarks of their respective owners.