Comparison table · Master data management

    Master Data Management Tools Comparison 2026

    An MDM tools comparison table: 24 master data management hubs and open-source entity resolution tools compared on 7 attributes, from how they match records and build golden records to where they run, what they cost, and a notable caveat. Plain text instead of checkmarks, since most differences between tools are not yes/no.

    Last updated: November 2026 · For every feature with sourced justifications, see the MDM Tool Feature Matrix · Integration Matrix · Master Data Management Tools List

    AttributeInformatica IDMC by SalesforceReltio by SAPSemarchyProfiseeStibo SystemsCluedInAtaccama ONEZinggSplinkAtroCoreLakeFusionTamrIBM Match 360Oracle Enterprise Data ManagementPrecisely EnterWorksBoomi DataHubSyncariVeratoPimcoreProspecta MDOSenzingAWS Entity ResolutionQuantexaTilores
    Core StrengthFull MDM suite with packaged 360 apps per domain (Customer, Supplier, Product, Reference 360)Real-time SaaS MDM with pretrained LLM matching and AI agentsRule-based MDM with a strong data model, on SaaS, self-hosted or as a Snowflake Native AppMicrosoft-first MDM: the only native MDM workload in Fabric, plus PurviewProduct data: PIM and multi-domain MDM in one platformAzure MDM with many built-in enrichers and an MCP serverMDM inside a data quality platform, with real-time APIs and Kafka eventsWarehouse-native ML entity resolution in Databricks, Fabric, AWS, GCP or SnowflakeFree probabilistic record linkage in Python, on DuckDB or SparkOpen-source MDM and PIM hub with a data model you configure yourselfDatabricks-native MDM that matches, merges and stewards golden records in your workspaceML entity resolution with curator review and persistent Tamr IDsMulti-domain MDM service on Cloud Pak for DataGoverns hierarchies, dimensions and reference data across Oracle appsMulti-domain MDM, PIM and DAM in one hubGolden records synced back to apps on the Boomi iPaaSAgentic MDM with built-in two-way sync to CRM, ERP and SnowflakeHealthcare identity MDM with referential matchingPIM-led MDM, DAM and DXP on one data modelWorkflow-driven master data governance for SAPEmbeddable entity resolution SDK, real-time and explainableManaged entity resolution on S3, priced per recordML entity resolution feeding graph analytics and investigationsReal-time entity resolution with an auditable entity graph
    Matching & Golden Records
    MatchingAI match and merge with rule recommendationsPretrained LLM matching plus fuzzy and exact rules, in real timeSemQL match rules with scores; duplicate check on entryML matching engine plus match rulesAI-powered matching (claimed); rules not documented publiclyFuzzy and phonetic rules in deduplication projectsRules, real-time and incremental; ML matching experimentalML model trained from 40–50 labeled pairs; rules in paid editionsFellegi-Sunter model estimated without labelsWeighted match rules in scheduled, incremental jobsRules, then a Random Forest, then an LLM for uncertain pairsML pair scoring with confidence labelsProbabilistic, deterministic or rule-based, ML-guidedNode matching and deduplication (method not public)Match, merge and survivorship rulesSimple-to-fuzzy matchingField rules with strictness, on every create or updateReferential matching against 300M+ identitiesBasic deduplication; no ML matchingRule-based duplicate checks on record creationBuilt-in principle-based rules; no model trainingRule-based, ML or data provider matchingML models, batch or at request timeExplicit rules plus ML scoring, as records arrive
    Golden RecordsSurvivorship, steward queues, hierarchiesField-level survivorship, steward review, unmergeSurvivorship, duplicate manager, history as of a dateSurvivorship, side-by-side steward review, audit trailSurvivorship, history, hierarchiesSurvivorship, steward review, history, hierarchiesMerge rules, matching proposals, historyNone yet: stewardship and golden records announcedNone: outputs clusters, golden records are up to youClusters, consolidation script, steward confirm/reject/splitSurvivorship, steward review, hierarchies, PIMConsolidation rules, curator queues, activity logMaster data entities that stewards reviewChange requests with approvals, auditsGolden records, audit trails, workflowsSurvivorship, quarantine workflowsField-level survivorship, written back to systemsSmart Views per workflow, steward task queueConsolidation and survivorship, audit trailNo golden records or survivorshipNo golden records: entities with representative featuresMatch groups with a match ID; no golden recordsResolved entities for network graphsBest value per field picked at query time
    Business
    Platform FitInformatica cloud; BigQuery extension, Databricks and Snowflake previewsPipelines to Snowflake, Databricks, BigQuerySnowflake Native App; Fabric in previewFabric native; connectors to Snowflake and DatabricksAzure-hosted SaaS or on-premisesAzure, Fabric (OneLake export), PurviewAtaccama Cloud or self-managedDatabricks, Fabric, AWS, GCP; Snowflake Native App in paid editionsDuckDB, Spark, PostgreSQL, SQLiteOwn database (PostgreSQL, MySQL, MariaDB); ERP and e-commerce connectorsDatabricks and Unity Catalog on AWS or AzureConnectors to Snowflake, BigQuery, OneLakeIBM Cloud Pak for Data and watsonxOracle Cloud EPM and FinancialsProduct-heavy catalogs; SAP onboardingBest inside an existing Boomi estateSalesforce, NetSuite, SAP, SnowflakeHealthcare: EHRs, HL7, FHIR; Snowflake appsPIM-led teams; GraphQL DatahubSAP ERP, Microsoft, CoupaRuns anywhere you run code: VMs, Docker, Kubernetes, air-gappedAWS only: S3, Glue, Clean RoomsBanks and public sector; Spark and Elastic underneathAWS-native, also self-hosted; GraphQL API
    Deployment ModelSaaSSaaSSaaS, self-hosted, Snowflake Native AppSaaS, self-hosted, hybridSaaS, self-hostedSaaS, self-hosted (your Azure subscription)SaaS, self-hostedSelf-hosted in your own cloud accountSelf-hosted Python librarySaaS, self-hostedSelf-hosted in your cloud account (PaaS)SaaS (Tamr Cloud), self-hosted (Tamr Core)SaaS, self-hosted (Cloud Pak for Data)SaaS (Oracle Cloud)Self-hosted (version 11.x)SaaS (Boomi platform)SaaS, or in your AWS or GCP accountSaaSSaaS (PaaS), self-hostedSaaS (multi-tenant)Self-hosted SDKAWS managed serviceSelf-hosted or cloudManaged AWS or self-hosted
    PricingQuote from salesQuote from salesQuote from salesQuote from sales; Azure MarketplaceQuote from salesPay-as-you-go SaaS on Azure; quote for private SaaSQuote from salesFree (AGPL-3.0); paid editions priced per deploymentFree (MIT)Free (GPL-3.0); SaaS from €400/month$16,667/month on AWS MarketplaceQuote from sales, priced per golden recordQuote from salesOracle Cloud subscriptionQuote from salesQuote from salesQuote from sales; free trialAnnual subscription, four tiers, quoteFree Community under €5M revenue; Enterprise $29,900/yrQuote from salesFrom $58,560/year (10M records)$0.25 per 1,000 recordsQuote from salesQuote by entity count; free local Studio
    Notable CaveatDocs need a login; Snowflake and Databricks features were still in preview in 2026SaaS only, no self-hostingNo ML matching; enrichment connectors are customer-builtDocs need a login, so many cells are unconfirmedStrongest on product data; matching details aren't publicNo ML matching; matching re-runs per projectML matching is experimental; hierarchy configuration is obsoleteAGPL-3.0; persistent IDs and incremental runs need a paid editionNo golden records, no persistent IDs, no stewardshipNo ML matching, no enrichment connectors, no warehouse write-backDocs portal is private; facts from release postsRule matching and real-time API are limited releaseMany features unconfirmed in public IBM docsFinance and planning dimensions, not customer recordsNo ML matching; SaaS and APIs not documented publiclyNo ML matching, hierarchies or reference data describedFixes data in source systems, no hub authoringHealthcare-only; no self-hostingPOCL license is not OSI open sourceNo survivorship, events or warehouse connectorsNo REST API, no UI for stewardsNo steward review; incremental runs not for ML matchingNo public docs; not a stewardship MDMGraphQL only; no steward review queue

    Distilled from our detailed, sourced feature-by-feature comparison. See the full matrix for justifications and links to primary sources.

    How I check every fact

    Checked against vendor docs

    I check each feature, integration and price against the vendor's docs, pricing page or GitHub repository, and link the fact to that source.

    Open to vendor corrections

    Vendors and users can send corrections. I verify them against the docs before changing anything, and every page shows when it was last updated.

    Spot a mistake? Send a correction →

    Frequently Asked Questions

    Which master data management tool has the most features?
    Out of 31 features in the MDM feature matrix, Profisee supports 24, LakeFusion supports 24, Informatica IDMC by Salesforce supports 23, CluedIn supports 19, Ataccama ONE supports 19. Full MDM hubs score high because they match, merge, store and serve golden records; entity resolution tools like Zingg and Splink only do the matching step and leave the rest to your data platform.
    Which MDM tools run inside Snowflake, Databricks or Microsoft Fabric?
    Inside Snowflake: Semarchy, Zingg, and Verato. Inside Databricks: Zingg, and LakeFusion. Inside Microsoft Fabric: Profisee, and Zingg. The other tools copy data into their own hub and publish golden records back through connectors.
    Which MDM tools use machine learning to match records?
    13 of the 24 tools document ML or probabilistic matching that scores record pairs: Informatica IDMC by Salesforce, Reltio by SAP, Profisee, Stibo Systems, Zingg, Splink, LakeFusion, Tamr, IBM Match 360, Verato, AWS Entity Resolution, Quantexa, and Tilores. The others match with rules you write, sometimes with AI that suggests rules.
    Which MDM tools keep a persistent entity ID?
    Reltio by SAP, Semarchy, Ataccama ONE, Zingg, AtroCore, Tamr, Senzing, AWS Entity Resolution, and Tilores document an ID per entity that stays stable when records merge or new data arrives, so downstream systems can store it as a key. Splink and the free edition of Zingg rebuild cluster IDs on every run.
    Which MDM tools have an MCP server?
    14 of the 24 tools ship an MCP server so AI agents can search and read golden records: Informatica IDMC by Salesforce, Reltio by SAP, Semarchy, Profisee, Stibo Systems, CluedIn, Ataccama ONE, LakeFusion, Tamr, Precisely EnterWorks, Syncari, Verato, Pimcore, and Senzing.
    How is this MDM tools comparison made?
    Each of the 24 tools is checked against 31 features using its public documentation and product pages. The feature matrix links every cell to its source; a dash means the public pages do not say either way, which happens most for vendors whose docs need a login. Vendors can request a correction through the feedback form.