vs

Manhattan vs ScyllaDB

Manhattan: Twitter's in-house, multi-tenant, eventually consistent key-value store with pluggable storage engines. ScyllaDB: C++ rewrite of Cassandra with shard-per-core design; also speaks the DynamoDB API.

key differences

  • Manhattan is primarily a key-value database; ScyllaDB is wide-column.
  • Licensing: Manhattan is proprietary (Proprietary (internal only)); ScyllaDB is source-available (ScyllaDB Software License Agreement 1.1).
  • Only ScyllaDB has secondary indexes, json, change data capture, materialized views, ttl / expiry.

choose manhattan for

  • +Reference design for multi-tenant, multi-datacenter key-value platforms
  • +Studying pluggable storage engines and replica reconciliation

best at · OLTP, Time-series

choose scylladb for

  • +Chat and messaging history at very high volume
  • +Low-latency feature stores for ML
  • +Cassandra or DynamoDB migration targets
  • +IoT and telemetry ingestion

best at · OLTP, Time-series, IoT

overview
Primary model
Key-Value
Wide-Column
Also supports
Time-Series
Key-Value
Type
Database engine
Database engine
Status
Active
Active
Developer
Twitter (now X Corp.)
ScyllaDB Inc.
First release
2012
2015
Latest version
–
2026.3.4 (Oct 2026)
License
ProprietaryProprietary (internal only)
Source-availableScyllaDB Software License Agreement 1.1
Free tier
—
✓
popularity
WhichDB Score
0.0
45.9
Rank
#775
#36
GitHub stars
–
15.8K
Stack Overflow questions
–
381
scale & guarantees
Comfortable data size
Massive (Multi-PB)
Massive (Multi-PB)
Horizontal scaling
Native
Native
Multi-region
Active active
Active active
ACID
Limited
Limited
Consistency
Eventual, Strong
Tunable, Eventual
Schema
Schemaless
Strict
Architecture
Distributed, Multi primary
Distributed, Multi primary, Client server
deployment
Self-hosted
—
✓
Vendor-managed cloud
—
✓
Serverless
—
—
Embedded
—
—
Edge
—
—
Clouds
–
AWS, Google Cloud
capabilities
Transactions
—
—
Joins
—
—
Secondary indexes
—
✓
Full-text search
—
—
Vector search
—
—
Geospatial
—
—
JSON
—
✓
Change data capture
—
✓
Triggers
—
—
Stored procedures
—
—
Materialized views
—
✓
TTL / expiry
—
✓
Pub/Sub
—
—
Columnar storage
—
—
In-memory
—
—
Durable persistence
✓
✓
Compression
✓
✓
Encryption at rest
—
✓
Role-based access
—
✓
Time travel
—
—
Branching
—
—
Online schema changes
—
✓
Streaming ingest
—
✓
Offline sync
—
—
Runs in WebAssembly
—
—
interfaces
Query languages
Key-value API
CQL, DynamoDB API
Client languages
–
Java, Python, Go, Rust, C++, C#, JavaScript, Scala
Best for
OLTP, Time-series
OLTP, Time-series, IoT, Feature store

Manhattan vs ScyllaDB: FAQ

What is the difference between Manhattan and ScyllaDB?

Manhattan: Twitter's in-house, multi-tenant, eventually consistent key-value store with pluggable storage engines. ScyllaDB: C++ rewrite of Cassandra with shard-per-core design; also speaks the DynamoDB API. Manhattan is primarily a key-value database; ScyllaDB is wide-column. Licensing: Manhattan is proprietary (Proprietary (internal only)); ScyllaDB is source-available (ScyllaDB Software License Agreement 1.1). Only ScyllaDB has secondary indexes, json, change data capture, materialized views, ttl / expiry.

When should I choose Manhattan over ScyllaDB?

Choose Manhattan for reference design for multi-tenant, multi-datacenter key-value platforms; and studying pluggable storage engines and replica reconciliation. Choose ScyllaDB for chat and messaging history at very high volume; low-latency feature stores for ML; and cassandra or DynamoDB migration targets.

Is Manhattan or ScyllaDB more popular?

ScyllaDB ranks #36 on the WhichDB Score (45.9/100) and Manhattan ranks #775 (0.0/100). Popularity is not fit: check the requirements that matter for your project.

Are Manhattan and ScyllaDB open source?

No. Manhattan is proprietary, closed-source software. Not in the OSI sense. ScyllaDB is source-available: you can read the code, but the license restricts some uses, typically offering it as a competing managed service. License: ScyllaDB Software License Agreement 1.1.