ELSER
Elastic's own sparse retrieval model, served on the Elastic Inference Service.
At a glance
Current release
From the release row ELSER v2. ELSER v2 on the Elastic Inference Service, Elastic Cloud Serverless, read with AWS us-east-1 selected. An embedding model: only input tokens are billed, so there is no output price.
Family record: Official documentation · 5 Sep 2026
Built on it
1 product in this catalogue names this family on their own pages.
Releases
-
ELSER v1
Shipped in Elasticsearch 8.8 as a technical preview: 'Elastic's proprietary transformer for semantic search'.
-
ELSER v2
Generally available in Elastic 8.11 in portable and x86-optimized builds; served on GPUs through the Elastic Inference Service from 2025-10-09 (technical preview, GA in Stack 9.2). ELSER v3 is described as in development.