The API Report CardAPI Index
Erwin

Erwin API

erwin.com

Erwin's primary API (SCAPI) is COM/IDispatch automation inside the Windows desktop app, covering the full model graph. REST exists only at the edges: Mart Server report endpoints with 4 hour bearer tokens, plus separate Evolve and Data Intelligence APIs.

Last verified: July 2026Software & Data Tools
API GRADE
D+
VERIFIED JUL 2026

SCORECARD

ExistencePOORThe primary API (SCAPI) is COM/IDispatch desktop automation, Windows only; no first class REST for the model graph.
AccessPOORRequires a licensed Data Modeler or Mart install; the trial edition caps at 25 entities, blocking real evaluation.
CoveragePOORThe full model graph is reachable only via Windows COM; the network facing REST surface is limited to Mart reports.
AuthPOORMart bearer tokens default to a 4 hour lifetime, and a known bug rejects valid tokens on Web Portal API calls.
Docs & DXGOODAPI Reference PDFs ship per release, the Mart Portal includes a Swagger UI, and Evolve has hosted apimatic docs.
StabilityMIXEDEach release keeps its own API reference and quirks; 2021 R1 regressed Mart connection stability.
MORE FROM THE REPORT CARD
Supergood: Erwin has an API, but using it means gates, contracts, or workarounds. Ours doesn't: stable endpoints, normalized JSON, managed auth.

Frequently asked questions

Erwin scores D+ on the API Report Card. Erwin's primary API (SCAPI) is COM/IDispatch automation inside the Windows desktop app, covering the full model graph. REST exists only at the edges: Mart Server report endpoints with 4 hour bearer tokens, plus separate Evolve and Data Intelligence APIs.

Tried to integrate with Erwin?
SOURCES
Primary API (SCAPI) is COM/IDispatch and runs inside the Windows desktop process, no first-class REST/HTTPS surface for the actual modeling object graph; non-Windows clients are effectively excluded bookshelf.erwin.com β†—
API documentation is sparse and version-fragmented across r7, r8, 2018 R1, 2019 R1, 2020 R2, 2021 R1, 12.0, 12.5, 13.x, 14.x, 15.x, every release lives at a separate bookshelf URL, and examples of successful code are scarce bookshelf.erwin.com β†—
Community-side API knowledge is thin, official forum threads note 'the API is very sparsely understood by the user community' and users routinely raise concerns about safely concurrently editing models in UI and via API support.erwin.com β†—
Mart Server REST bearer tokens default to a 4-hour lifetime, requiring re-authentication for long-running integrations; lifetime is tunable but only via the Mart Portal Advanced configuration UI, not programmatically bookshelf.erwin.com β†—
Known Mart authentication bug, token returned from auth call is not always passed to subsequent Web Portal API requests, producing 'A valid token must be supplied. Call the login API first to obtain the token' errors support.erwin.com β†—
Mart connection stability regression, in 2021 R1, opening a second model from Mart within a few minutes produces access-denied errors, whereas earlier 9.7 sessions held connections for hours peerspot.com β†—
Multiple parallel API surfaces (SCAPI COM, Mart Server REST, erwin Evolve REST, erwin Data Intelligence REST) each have different idioms, auth flows, and lifecycles, integrators must learn three or four distinct APIs depending on which product they touch support.quest.com β†—
Mart Server REST API is functionally scoped largely to running **predefined** reports, building genuinely new programmatic queries against the model graph requires either authoring custom Mart reports or dropping back to SCAPI on the desktop bookshelf.erwin.com β†—
Mart Server is typically deployed on-prem or in a customer-managed cloud VM behind a corporate VPN/firewall, which blocks third-party SaaS integrations from reaching the API without customer-side network work support.quest.com β†—
Endpoint paths, payload shapes, and Mart database schema change across annual releases (2018 R1 β†’ 2019 R1 β†’ 2020 R2 β†’ 2021 R1 β†’ 12.x β†’ 13.x β†’ 14.x β†’ 15.x), requiring re-validation and Mart database upgrades on each customer upgrade support.quest.com β†—
Third-party community wrappers (e.g. SSA's erwin-api-wrapper .NET library) exist precisely because the raw COM surface is awkward to consume directly, and Quest does not publish a first-party modern SDK in Python / JavaScript / Go github.com β†—
No first-party public developer sandbox or hosted test instance, evaluating the API requires installing the Windows fat-client (with a 25-entity trial cap) or standing up a full Mart Server, both gated behind Quest sales for production licensing quest.com β†—
UI and UX have not aged well, interface is described as clunky, dated, dense, and slow, with steep onboarding for new users g2.com β†—
Performance degrades materially on large models, generating reports and PDFs of big diagrams is slow and the application has been reported to crash trustradius.com β†—
Cost / licensing is widely cited as the #1 barrier to entry; Standard ~$200-299/mo and Workgroup ~$399/mo per named user with mandatory annual maintenance, and enterprise deals require a quote peerspot.com β†—
Trial edition is artificially capped at 25 entities, which makes it nearly impossible to evaluate the tool against a real enterprise schema before purchase capterra.com β†—
Reviewers consistently say the product 'stagnated a long time ago', limited cloud-native support, no native Linux/macOS client, weak big-data support, Oracle reverse-engineering issues thectoclub.com β†—
Collaboration is weak without Mart Server; Mart adds a substantial deployment burden (Tomcat + PostgreSQL + license) just to get multi-user version control g2.com β†—
Reporting is cumbersome, users routinely export to Excel/HTML and manipulate externally because in-product reporting falls short trustradius.com β†—
erwin Data Catalog / Data Intelligence value depends heavily on the quality of upstream metadata; with poor source metadata the catalog benefit is limited gartner.com β†—
Windows-only desktop client is increasingly a friction point for Mac-and-Linux-first modern data teams; running erwin DM via Citrix/VDI is a common workaround peerspot.com β†—
Cloud integration capabilities lag cloud-native competitors (SqlDBM, Hackolade, dbdiagram.io); 14.x/15.x have added dbt, git and Fabric Lakehouse support but the architecture remains desktop-first erstudio.com β†—
Total cost of ownership compounds across Data Modeler, Mart Server, Data Intelligence, Data Catalog and Evolve modules, customers commonly need a Quest sales engagement just to scope a complete deployment myriadinc.net β†—
Internal portfolio overlap with Toad Data Modeler (also a Quest product) creates customer confusion about which tool to standardize on quest.com β†—