Model 204

active

by Rocket Software · since 1972 · also known as M204, Rocket Model 204

Rocket Software's mainframe inverted-list DBMS, in production since 1972 and still getting releases.

multivalue (pick)database engineproprietarylegacy

Overview

Model 204 is a database management system for IBM z/OS mainframes, built by Computer Corporation of America, first deployed in 1972 and now owned and developed by Rocket Software. It predates the relational model: files are inverted-list structures that combine its own bitmap indexes with hashing, B-trees and record lists, applications are written in its SOUL (User Language) 4GL, the Janus add-ons provide web, socket and TLS services, and Model 204 SQL Server gives limited SQL access through explicit file mappings. It copes with very large record counts and heavy transaction rates, which is why agencies such as Services Australia still run core systems on it after a replacement project was scrapped in 2023. Version 8.0 (2026) added TLS 1.3, a built-in SSH server and hardware-accelerated compression: alive, but a system you inherit rather than choose.

Strengths & trade-offs

Strengths

  • +Decades of proven operation for very large government and defense workloads
  • +Bitmap, hash and B-tree indexing gives fast retrieval on huge files
  • +Integrated 4GL (SOUL) plus web and socket serving via Janus
  • +Still actively released by Rocket (8.0 with TLS 1.3 and SSH)

Trade-offs

  • −Mainframe-only (z/OS) and expensive to operate
  • −Pre-relational: SQL access is limited and needs explicit mapping
  • −Shrinking pool of SOUL/User Language developers
  • −Migrations off it are notoriously hard and slow

Great for

  • +Maintaining existing government and defense mainframe systems
  • +High-volume record retrieval and update on z/OS
  • +Web-enabling legacy M204 applications through Janus

Think twice for

  • −New application development
  • −Cloud-native or distributed deployments
  • −Teams without mainframe skills

Capabilities

Native or official first-party support · 5 of 25 tracked

✓Transactions
✓Secondary indexes
✓Stored procedures
✓Durable persistence
✓Role-based access
–Joins
–Full-text search
–Vector search
–Geospatial
–JSON
–Change data capture
–Triggers
–Materialized views
–TTL / expiry
–Pub/Sub
–Columnar storage
–In-memory
–Compression
–Encryption at rest
–Time travel
–Branching
–Online schema changes
–Streaming ingest
–Offline sync
–Runs in WebAssembly

Scale & consistency

Comfortable data size
Medium · ≤ 10 TB
Horizontal scaling
None
Multi-region
No
Transactions
Limited ACID
Consistency
Strong
Schema
Flexible
Architecture
Single node, Client server
Query languages
SOUL (User Language), SQL (via Model 204 SQL Server)

Scales vertically on z/OS; known for very large files and high transaction rates on a single system.

Deployment

Self-hosted Vendor-managed cloud Serverless Embedded Edge Kubernetes operator

Ecosystem

apis & protocolsSOUL, Janus Web Server, Janus Sockets, Connect*
client languagesSOUL
used byServices Australia

Alternatives to Model 204

Compare all→
Software AG AdabasMultivalue (Pick)8.4

Inverted-list, multi-value mainframe database paired with Natural, run as a standalone Software AG business.

model-204 vs adabas →
IBM IMSHierarchical7.7

IBM's 1960s hierarchical database and transaction manager, still running bank and airline cores on z/OS.

model-204 vs ibm-ims →
IDMSNetwork / Navigational6.2

CODASYL network-model mainframe database, now Broadcom's, still running decades-old COBOL systems.

model-204 vs idms →
IBM Db2Relational45.5

IBM's enterprise RDBMS for Linux, UNIX, Windows and z/OS, with BLU columnar and pureScale clustering.

model-204 vs ibm-db2 →
Rocket UniData / UniVerseMultivalue (Pick)11.0

Rocket's Pick-style MultiValue databases, powering thousands of vertical business applications.

model-204 vs rocket-unidata-universe →
jBASEMultivalue (Pick)1.4

MultiValue database that compiles Pick BASIC to native code, now part of Rocket's MV portfolio.

model-204 vs jbase →

Model 204 FAQ

What is Model 204?

Model 204 is a multivalue (Pick) database by Rocket Software, first released in 1972. Model 204 is a database management system for IBM z/OS mainframes, built by Computer Corporation of America, first deployed in 1972 and now owned and developed by Rocket Software. It predates the relational model: files are inverted-list structures that combine its own bitmap indexes with hashing, B-trees and record lists, applications are written in its SOUL (User Language) 4GL, the Janus add-ons provide web, socket and TLS services, and Model 204 SQL Server gives limited SQL access through explicit file mappings. It copes with very large record counts and heavy transaction rates, which is why agencies such as Services Australia still run core systems on it after a replacement project was scrapped in 2023. Version 8.0 (2026) added TLS 1.3, a built-in SSH server and hardware-accelerated compression: alive, but a system you inherit rather than choose.

Is Model 204 open source?

No. Model 204 is proprietary, closed-source software. License: Commercial (Rocket Software).

Does Model 204 support ACID transactions?

Partly. Model 204 offers limited ACID guarantees (for example, single-record or single-partition transactions), with strong consistency.

Does Model 204 scale horizontally?

Model 204 does not scale horizontally; it runs on a single node, so you scale it up rather than out. It is comfortable at medium scale (≤ 10 TB).

How can I run Model 204?

You can self-host it.

What is Model 204 best for?

Model 204 is a strong fit for maintaining existing government and defense mainframe systems; high-volume record retrieval and update on z/OS; and web-enabling legacy M204 applications through Janus. Think twice for new application development.

What are the best alternatives to Model 204?

Common alternatives to Model 204 include Software AG Adabas, IBM IMS, IDMS, IBM Db2 and Rocket UniData / UniVerse. The right choice depends on your workload, scale and hosting; WhichDB's smart match ranks them against your requirements.

How popular is Model 204?

Model 204 ranks #407 of 844 databases on the WhichDB Score (4.4/100), and #3 among multivalue (Pick) databases. The score combines public adoption signals such as package downloads, GitHub activity, job posts and developer discussion.