OpenSearch

OpenSearch is an Apache 2.0-licensed search and analytics engine created in 2021 when AWS forked Elasticsearch after Elastic moved to proprietary licensing.

Reviewed by 7wData

On this page

Publisher review

OpenSearch is an Apache 2.0-licensed search and analytics engine created in 2021 when AWS forked Elasticsearch after Elastic moved to proprietary licensing. It includes a REST API, OpenSearch Dashboards (the visualization layer), and built-in enterprise security—LDAP, SAML, field-level access control, and audit logging—all free. As a fully open-source project now governed by the OpenSearch Software Foundation (a Linux Foundation initiative), it's positioned as the lowest-cost, least-locked-in alternative to Elasticsearch.

On AWS OpenSearch Service, costs run 30–50% lower than Elasticsearch Cloud on comparable workloads, with on-demand instances starting at $0.036/hour for small clusters and cold storage at $0.01/GB-month. OpenSearch supports hybrid keyword and vector search (16,000-dimensional vectors via HNSW and Faiss, with binary quantization for 32x compression), making it appealing for semantic search and observability at scale. Real-world deployments include Uber (billion-scale vector search), SAP (large observability infrastructure), Coursera, and Fidelity.

However, OpenSearch lags Elasticsearch on observability features—missing query autocomplete (in development since 2019), Log Stream views, wildcard types, and auto-refresh. Community criticism centers on development velocity: key features remain unfinished while Elasticsearch pushes ahead, and support is AWS-centric. The project is backward-compatible with Elasticsearch 7.x APIs but not forward-compatible with Elasticsearch 8+, creating migration friction for teams on older stacks.

Best suited for teams migrating from Elasticsearch seeking cost reduction and open licensing, or those building vector search without cloud vendor lock-in. For pure log analytics, ClickHouse often wins on compression; for simpler full-text search, Meilisearch or Typesense may be faster to deploy.

Get the AI & data signal, daily.

48k+ subscribers read this every morning. One email, both newsletters. Unsubscribe anytime.

How it works

  1. Hybrid keyword + vector search

    Supports both full-text search and dense vector retrieval (up to 16,000 dimensions) in a single query, using Lucene HNSW or Faiss indexing with optional binary quantization.

  2. Enterprise security included

    Field-level access control, LDAP/SAML authentication, role-based access, and audit logging built in—all free, where Elasticsearch requires paid premium tiers.

  3. OpenSearch Dashboards visualization

    Browser-based UI for building dashboards, running queries, and monitoring indices; code-driven and customizable, with built-in anomaly detection and alerting.

  4. Apache 2.0 open-source license

    Fully open-source with no restrictions on commercial use, redistribution, or modifications; unlike Elasticsearch's proprietary SSPL/ELv2 licenses.

  5. Managed service on AWS (pay-per-use)

    AWS OpenSearch Service offers on-demand and reserved instances, serverless OCUs, and tiered storage (hot/warm/cold) with transparent pricing from $0.036/hour.

  6. Backward-compatible REST API

    Wire-compatible with Elasticsearch 7.x client libraries and APIs, allowing teams to migrate without rewriting all application code.

  7. Anomaly detection and time-series analytics

    Built-in tools for detecting statistical outliers in streaming data and running aggregations at scale for monitoring and observability workloads.

Strengths and trade-offs

Strengths

  • 30–50% lower total cost of ownership than Elasticsearch Cloud on AWS, with enterprise security features free rather than paywalled.
  • Fully open-source Apache 2.0 license with no vendor lock-in; governed by Linux Foundation, not a single company.
  • Mature hybrid search combining full-text and vector retrieval (up to 16,000 dims with compression) in one system, appealing for semantic and observability use cases.

Trade-offs

  • Significant feature lag behind Elasticsearch on observability (missing query autocomplete, Log Stream, wildcard type, auto-refresh) with slow progress on core requests.
  • AWS-centric ecosystem; managed offerings primarily on AWS, with limited native support on Azure or GCP, creating cloud lock-in despite open licensing.
  • Not forward-compatible with Elasticsearch 8+; Elasticsearch 7.x APIs only, creating friction for teams on modern stacks and reducing long-term flexibility.

Pricing context

OpenSearch itself is free under Apache 2.0. AWS OpenSearch Service (the dominant managed option) uses on-demand pricing: small instances (t3.small) start at $0.036/hour, medium clusters (m6g.large) at $0.128/hour, with storage at $0.08/GB-month for hot (gp3), $0.03 for warm (UltraWarm), and $0.01 for cold. Serverless costs $0.248/OCU-hour (roughly 6 GB RAM).

Reserved instances offer 30–60% discounts. Total cost on AWS typically 30–50% lower than Elasticsearch Cloud for equivalent workloads because enterprise security (LDAP, SAML, field-level access) is included free, not a premium add-on. Self-managed deployments have zero licensing cost; only infrastructure expenses apply.

Alternatives

User reviews

No user reviews yet. Be the first to write one.

Sources

Reporting on this tool draws on these publicly available sources.

  1. opensearch.org — Official OpenSearch project homepage, governance structure, and Apache 2.0 license confirmation.
  2. docs.opensearch.org — Core feature documentation, vector search (HNSW/Faiss), security (LDAP/SAML), anomaly detection, and breaking changes.
  3. docs.opensearch.org — Migration guide from Elasticsearch 7.x, backward compatibility statement, and forward incompatibility with Elasticsearch 8+.
  4. aws.amazon.com — AWS OpenSearch Service pricing: on-demand instances (t3.small $0.036/hr), storage tiers, reserved instances, and cost examples.
  5. www.uber.com — Real-world adoption case: Uber's billion-scale vector semantic search deployment and architecture on OpenSearch.
  6. www.infoworld.com — Overview of OpenSearch in 2025, differentiation from Elasticsearch, community momentum, and adoption trends.
  7. bigdataboutique.com — Cost comparison (30–50% savings), feature lag analysis (query autocomplete, wildcard type, Log Stream), and performance trade-offs.
  8. news.ycombinator.com — Community criticism: feature stagnation, query autocomplete half-decade in development, support limitations, and honest developer feedback.
  9. signoz.io — Technical feature comparison, observability gaps, API compatibility, vector search capabilities, and licensing implications.
  10. coralogix.com — AWS OpenSearch cost breakdown, hidden costs (ingestion, cross-region transfer), serverless OCU pricing, and comparison to Elasticsearch.