- Home
- AI Search and Discovery Platforms Software
- Amazon Kendra
Amazon Kendra
What it is
Amazon Kendra is a cloud enterprise search service for websites and applications that uses machine learning to answer natural-language questions. It can match curated FAQs and extract answers from tables in HTML pages. The Kendra Retriever API returns relevant passages for retrieval-augmented generation, with ACL filtering, metadata filtering, and relevance tuning. Native connectors include Amazon S3, Microsoft SharePoint, Salesforce, ServiceNow, Google Drive, and Confluence, plus custom and partner connectors. Experience Builder creates and launches search applications without coding or machine-learning experience, with AWS IAM Identity Center, Azure AD, and Okta sign-on support. Search Analytics reports quality, usability, interaction, and content gaps through an API. Custom Document Enrichment can add metadata, convert scans to text, classify files, extract entities, and transform content. Kendra encrypts data in transit and at rest. Access is through the console, CLI, or SDKs. Plans start at the listed rates, with a 30-day trial.
For and against
For
- Natural-language answers and FAQ matching
- Connectors for major enterprise content systems
- No-code Experience Builder
- ACL and metadata filtering for retrieval
- Encryption in transit and at rest
Against
- No free plan
- Single ingested documents are limited to 50 MB
- GenAI Enterprise indexes are limited to four regions
- Components of a campaign must share one AWS Region
The verdict
Amazon Kendra combines machine-learning search, retrieval controls, connectors, and no-code application building. Its regional and document-size limits should be considered alongside the paid plans and available trial.
Who it suits
Amazon Kendra suits organizations building searchable websites or applications over connected enterprise content. It fits teams needing natural-language answers, permissions filtering, analytics, and document enrichment.
Plans and prices
GenAI Enterprise Edition
up to 20,000 documents or 200MB extracted text · 0.1 QPS (approximately 8,000 queries/day)
Read 1 Oct 2026 · aws.amazon.com
Basic Developer Edition
up to 10,000 documents or 3GB extracted text · 0.05 QPS (approximately 4,000 queries/day) · not recommended for production workloads
Read 1 Oct 2026 · aws.amazon.com
Basic Enterprise Edition
up to 100,000 documents or 30GB extracted text · 0.1 QPS (approximately 8,000 queries/day)
Read 1 Oct 2026 · aws.amazon.com
Connectors for GenAI Enterprise Edition
includes up to 500 hours of sync usage per month
Read 1 Oct 2026 · aws.amazon.com
On the record
- Access methods
- Kendra can be accessed through the Amazon Kendra console, AWS CLI, and AWS SDKs, and components of a campaign must be created in the same AWS Region.docs.aws.amazon.com · read 1 Oct 2026
- Compliance
- Amazon Kendra is compliant with HIPAA, SOC 2, IRAP, FedRAMP Moderate in US East and US West, and FedRAMP High in AWS GovCloud (US-West).docs.aws.amazon.com · read 1 Oct 2026
- Connectors
- Native connectors include Amazon S3, Microsoft SharePoint, Salesforce, ServiceNow, Google Drive, and Confluence, with custom and partner connectors also supported.aws.amazon.com · read 1 Oct 2026
- Document enrichment
- Custom Document Enrichment can add metadata, convert scanned documents to text, classify documents, extract entities, and transform content using rules or AWS Lambda.aws.amazon.com · read 1 Oct 2026
- Document-size limit
- The maximum size of a single document or raw file ingested through a data-source connector is 50 MB.docs.aws.amazon.com · read 1 Oct 2026
- Domain optimization
- Kendra is optimized for domains including IT, financial services, insurance, pharmaceuticals, legal, healthcare, telecommunications, automotive, and others.aws.amazon.com · read 1 Oct 2026
- Generative AI retrieval
- The Kendra Retriever API returns semantically relevant passages with ACL filtering, metadata filtering, and relevance tuning for retrieval-augmented generation workflows.aws.amazon.com · read 1 Oct 2026
- Identity integration
- Experience Builder integrates with AWS IAM Identity Center and supports identity providers such as Azure AD and Okta for single sign-on.aws.amazon.com · read 1 Oct 2026
- Natural-language search
- Kendra answers natural-language questions, matches curated FAQs, and can extract answers from tables in HTML pages.aws.amazon.com · read 1 Oct 2026
- No-code builder
- Experience Builder lets users create, customize, and launch a search application without coding or machine-learning experience.aws.amazon.com · read 1 Oct 2026
- Regional limit
- GenAI Enterprise Edition indexes are available only in US East (N. Virginia), US West (Oregon), Europe (Ireland), and Asia Pacific (Sydney).docs.aws.amazon.com · read 1 Oct 2026
- Search analytics
- The Search Analytics Dashboard reports search quality, usability, user interaction, and content-gap metrics and exposes the data through an API.aws.amazon.com · read 1 Oct 2026
- Security
- Kendra encrypts data in transit and at rest and supports AWS-owned, AWS-managed, or customer-managed KMS keys for at-rest encryption.aws.amazon.com · read 1 Oct 2026
- What it does
- Amazon Kendra is an intelligent enterprise search service powered by machine learning for websites and applications.aws.amazon.com · read 1 Oct 2026