Skip to content
Tool Review · NoSQL & Distributed Databases

RavenDB

RavenDB is RavenDB Ltd's transactional document database with its own RQL query language, automatic indexes, sharding, time series, counters and ETL, available with free Community and Developer licences, paid per-core licences by quote, or as RavenDB Cloud.

Freemium Research-based review
Last verified November 2026. Version checked: 7.2.7 (stable download, November 2026). Features and pricing for commercial products change; check the vendor site for the latest details.
Vendor
RavenDB Ltd
Price
Freemium
Platforms
Windows, Linux, macOS, Docker, Raspberry Pi, RavenDB Cloud on AWS, Azure and Google Cloud
Licence
Source under AGPLv3; servers run under RavenDB licences (free Community and Developer, paid Professional, Enterprise and AI)
Product type
Commercial document database (AGPLv3 source)
Official site
Works with
.NET (C#)JavaNode.jsPythonGoPHPRubyC++Apache Kafka (ETL)SQL databases (SQL ETL)

At a glance

Data modelDocument (JSON), with time series, counters and attachments
Query languageRQL (RavenDB Query Language)
Consistency modelNot documented as a single model on the pages reviewed; cluster-wide transactions available
ReplicationCluster replication with failover; external, hub-sink and filtered replication
Partitioning or shardingSharding splits a database across servers
TransactionsTransactions, including cluster-wide transactions across nodes
Deployment and cloud availabilityWindows, Linux, macOS, Docker, Raspberry Pi; RavenDB Cloud on AWS, Azure and Google Cloud
LicensingAGPLv3 source; free Community and Developer licences; paid Professional, Enterprise and AI by quote
Short answer

RavenDB suits teams, especially in .NET, who want a document database with transactions, automatic indexing and many built-in extras (time series, counters, ETL, vector search) in one server. The free Community licence is small (3 cores, 6 GB RAM) and paid licence prices are not published, so get a quote early if you expect to grow.

How we know: This profile is research-based: licence types and limits, platforms, features, the latest version and Cloud billing rules were checked on 7 November 2026 against ravendb.net (buy, features, download and cloud pages), the RavenDB 7.2 licensing documentation, the RavenDB Cloud billing documentation and the ravendb/ravendb repository. We have not run RavenDB.

Overview

RavenDB stores JSON documents and queries them with RQL, the RavenDB Query Language, which supports full-text search, facets, spatial queries and patching. Queries without a matching index trigger automatic indexes that adapt over time, and static indexes are available for heavier work. A server can run alone or in a cluster with failover and load balancing, cluster-wide transactions, sharding of a single database across servers, and external, hub-sink and filtered replication. Built-in time series, counters, revisions, subscriptions, attachments and document expiration cover needs that would otherwise require extra products.

RavenDB Ltd publishes the source on GitHub under AGPLv3 and ships clients for C#, Java, Node.js, Python, Go, PHP, Ruby and C++. ETL tasks push data to SQL databases, other RavenDB databases, OLAP storage, Elasticsearch, Kafka, RabbitMQ, AWS SQS, Azure Queue Storage and Snowflake. Recent releases add AI features: vector search, embedding generation through external services, generative AI tasks and an AI agent creator, sold with the AI licence. RavenDB Cloud runs managed clusters on AWS, Azure and Google Cloud.

Key features

  • JSON document store with cluster-wide transactions
  • RQL query language with full-text search, facets, spatial queries and patching
  • Automatic indexes created from queries, plus static indexes and rolling index deployment
  • Sharding of a database across servers, and external, hub-sink and filtered replication
  • Built-in time series, counters, revisions, subscriptions and attachments
  • ETL to SQL, OLAP, Elasticsearch, Kafka, RabbitMQ, AWS SQS, Azure Queue Storage and Snowflake
  • Vector search and generative AI integration (AI licence)

Integrations and ecosystem

.NETJavaNode.jsPythonGoApache KafkaRabbitMQElasticsearchSnowflakePower BIGrafanaOpenTelemetry

Pros and cons

Pros

  • Many capabilities in one server: documents, time series, counters, ETL and vector search
  • Automatic indexing reduces up-front index design
  • Free Community licence permits commercial production use
  • Clients for eight languages, with a long history in .NET

Cons

  • Community licence is limited to 3 cores, 6 GB RAM and 3 nodes, and must run the latest major version
  • No published prices for Professional, Enterprise, AI or RavenDB Cloud in the page text
  • RQL is specific to RavenDB, so skills and queries do not transfer directly
  • Smaller ecosystem and hosting choice than MongoDB

Pricing

Every RavenDB server needs a licence key. Two are free: Community, valid for commercial production use for one year at a time, limited to 3 cores and 6 GB of RAM across a cluster of up to 3 nodes (and up to 60 cores per customer entity), with only 2 document revisions kept and the requirement to run the latest major version; and Developer (9 cores, 36 GB RAM, 3 nodes), for development, CI and QA only, renewed every 6 months. Professional (6 GB of RAM per licensed core), Enterprise and AI (12 GB per core) are yearly per-core subscriptions, but the buy page publishes no prices: contact sales. RavenDB Cloud is billed by the minute (one-hour minimum) on demand, with 5% or 10% discounts for yearly production contracts; the documentation says prices depend on cloud, region and contract and are shown in the portal when you provision. We could not read the Cloud price list.

Pricing checked on the vendor's official page on 7 November 2026. Prices change; confirm before buying.

Who it is best for

Application teams, often on .NET, who want a transactional document database with automatic indexing and built-in time series, ETL and AI features.

Typical use cases

  • Line-of-business and SaaS applications storing aggregates as documents
  • IoT and monitoring data using built-in time series with rollups and retention
  • Feeding SQL databases, data lakes or message queues through built-in ETL
  • Semantic search over application data with vector search

Who should look elsewhere

  • Teams that need wide hosting and tooling choice for a document database, where MongoDB is more widely supported.
  • Budget-sensitive production systems that will outgrow 3 cores or 6 GB RAM and need a published price list.

Verdict

Our take

RavenDB packs an unusually broad feature set into one document database, with transactions and automatic indexing that ease development. Its free tier is generous enough to start and even run small production systems, but the per-core paid licences and Cloud prices are by quote, so compare total cost with MongoDB or Couchbase before scaling.

Frequently asked questions

Is RavenDB free?

There is a free Community licence for commercial production use, limited to 3 cores, 6 GB of RAM and a 3-node cluster, renewed yearly and tied to the latest major version, plus a free Developer licence for non-production use. Larger deployments need a paid Professional, Enterprise or AI licence.

How much does a RavenDB licence cost?

Paid licences are yearly subscriptions per core, with 6 GB of RAM per core for Professional and 12 GB for Enterprise and AI, but the buy page asks you to contact sales rather than listing prices.

Is RavenDB open source?

Its source code is published on GitHub under AGPLv3, but running a server requires a RavenDB licence key, including the free Community or Developer licence.

Which clouds does RavenDB Cloud run on?

AWS, Microsoft Azure and Google Cloud. It is billed by the minute with a one-hour minimum, with discounts for yearly production contracts.

Sources

Checked November 2026.

Listed in: NoSQL & Distributed Databases.

How we research reviews and what star ratings mean: our editorial method.

Compare your options

Browse the full tools directory or see the head to head comparisons.