This wiki has undergone a migration to Confluence found Here

Publicly Available FHIR Servers for testing

From HL7Wiki
Revision as of 20:06, 5 June 2015 by Jenni syed (talk | contribs) (→‎List: update FHIR sandbox endpoint)
Jump to navigation Jump to search

Back to FHIR home page


This page lists FHIR servers that are publically available for testing. In order to avoid spam etc, the servers are generally password protected. A contact is provided to get a password.


Note that these servers are testing servers. They may be sporadically unavailable, and as the FHIR specification is a moving target, they may not always implement the latest version, or do so correctly.

  • - Ewout's test server (previously The actual service endpoint is at
    • Supports all resource types, all operations, xml + json
    • implementation details: C# reference implementation, WebApi 2.0 library, Mongo DB for storage and search.
    • Server is running on AppHarbor, Mongo at MongoLab, and storage of binary is done on Amazon S3
    • Available in the open source - see [[2]]
  • - AEGIS WildFHIR - DSTU2 (2015May)
    • Recently updated to support the DSTU-2 (2015May) current ballot version of FHIR
      • NOTE: Update to DSTU-2 (2015May) required a reset/removal of all existing resource data
    • Supports all resource types and operations (except transaction); better support for search; support for conditional read, create/update return preference
    • Based on the FHIR DSTU-2 (2015May) Java reference implementation; hosted on RedHat WildFly 8.1.0
    • Test client interface:
  • - HealthConnex - DSTU2 (2015May)
    • DSTU-2 (2015May) current ballot version of FHIR
    • Supports all resource types, but not all operations
    • .NET (C#) implementation on SQL Server Azure
    • No support for Tags/search chaining or profiles at present
  • GNU Health FHIR server
    • Supports read, validate and search for Patient, DiagnosticReport, Practitioner, Procedure, Observation, Condition, FamilyHistory
    • DSTU1
    • A Flask app. It's connected to the GNU Health community server database. Consequently, it's possible to create and update patients, doctors, etc. through the GNU Health frontend and the changes should be reflected on the FHIR server (indirect write support, I suppose).
    • I think our goal is to use the FHIR server as an adapter for non-GNU Health EHRs and users, since we already have synchronization between instances. But, there are other interesting possibilites, too.
    • I'm updating the code frequently and it's still in the dev stages (e.g., adding new resources frequently).
    • Documentation is here: