Illustration of cloud, data cube, code panel, and bar chart connected by network lines on a light background

Introducing IBM Db2 Serverless Technical Preview: A serverless database for developers and agents

Spin up a Db2 database in seconds. Connect your application. Start building. No capacity to plan, no infrastructure to size, no operations to run.

We’re excited to announce the technical preview of IBM Db2 Serverless, a fully managed, cloud-native database service that makes it easy to create, connect, and scale databases without managing infrastructure. It brings the IBM Db2 engine enterprises already trust to a fully serverless model, with self-service provisioning, enterprise-grade security, and elastic scale that requires no capacity planning in advance.

As more applications and AI agents create and operate databases on their own, they need a database that responds at their speed and does not make them wait to provision or scale. Db2 Serverless meets that need without forcing a trade-off. Agents and developers get serverless speed on an engine proven across demanding, mission-critical workloads, with the security, data integrity and reliability those workloads require. Agents create and manage databases in code through REST and MCP interfaces, and snapshot-based branching provides isolated copies of production for development and testing.

Ready for an agent-first future

AI assistants and software agents now generate code, provision services and wire applications together. To join that loop, a database must be easy to create and manage programmatically.

Db2 Serverless ships the building blocks in the Tech Preview:

  • REST APIs for service management and provisioning
  • MCP Gateway support for service lifecycle operations
  • Automatic scaling and IBM-managed operations

This makes Db2 easy for developers and their tools to create, configure and manage in code, so standing up a database fits right into an agent-driven workflow.

Branch like you branch code

Snapshot-based branching is a standout at launch. You can now create isolated copies of a database for development and testing, without touching production. Test schema changes, validate updates, try new features and run parallel work, each in its own environment.

As AI-assisted development grows, branching matters more. Developers and agents need safe environments to experiment and iterate. Branching gives them that, and keeps production protected.

Scale to demand, pay for usage

Db2 Serverless cuts operational overhead and ties cost to actual use, so you never overprovision and never wait:

  • Elastic scaling: Compute follows workload demand, with no manual capacity planning
  • Scale-to-zero: Idle applications pause compute
  • Pay-as-you-go: Pay for what you use, start small, grow as you go
  • Automated maintenance: IBM handles infrastructure and engine updates with minimal disruption

Built on the Db2 engine you trust

Serverless simplicity with the Db2 foundation. The Tech Preview includes:

  • Standard SQL with full Db2 compatibility
  • OLTP for transactional applications
  • IBM-managed service resiliency for business continuity
  • Automated infrastructure and database maintenance
  • A production-ready service architecture

You get a production-grade engine that’s ready to build on from day one, so any team can build on it with confidence.

Available now in Technical Preview

IBM Db2 Serverless is a strong fit for transactional apps, new development, dev and test environments, proofs of concept and light analytics. The Tech Preview is your first chance to build on it. 

Join the waitlist to get early access

Author

Jillian Quiller

Growth Product Manager, Db2