A data catalog inventories what data an organization has, tables, dashboards, files, models, and records what each one means, who owns it and where it came from. Data governance software layers policy on top of that inventory: access rules, stewardship workflows and an audit trail of who touched what.
This guide ranks the platforms on what actually separates them once the demo is over: what a real rollout costs once metadata volume and seats are added, how much of the catalog you can get back out if you leave, and who is really allowed to hold the index of your company's data.
Vendors can pay for more visibility on this website, for example a higher place in a ranking or more room on a page. Commercial relationships can also affect which products we feature and how prominently. Every vendor link is marked the same, whatever the arrangement behind it. How that works.
In short
What data catalog software does
Data catalog software inventories an organization's data assets, tables, dashboards, files and models, and records what each one means, who owns it and where it came from. Data governance software layers policy on top of that inventory: access rules, stewardship workflows and an audit trail, and increasingly, the governed context an AI agent reads before it answers a question about the business.
We ranked the 11 data catalog tools on setup effort first, then total cost, then data export, then exit, then fit for the buyer. One rule overrides the order: smaller vendors take the first three places when the category offers them, and the best-known brands start at four. The reasoning behind the five tests, and behind having no score out of ten, is set out on the how we work page.
Free tier covers browse and search of Azure and Fabric sources up to 1,000 annotated assets; enterprise is consumption-based, billed per governed asset per day plus Data Governance Processing Units for data-quality rules, with published pay-as-you-go rates
Yes
Microsoft's business-facing data catalog, layered over the Purview Data Map metadata graph
Leipzig-built enterprise knowledge graph for governed, machine-readable metadata
Ranked #1 of 11 in Best Data Catalog Software in 2026.
Published pricingEurope
eccenca's Corporate Memory models an organisation's data and business rules as a semantic knowledge graph rather than a flat catalog table, with a stated goal of making the result explainable enough for AI systems built on top of it to be trusted.
Low-code tooling is aimed at subject-matter experts rather than only data engineers, which is a genuine differentiator from the more developer-first tools on this list. It is a smaller, German-only-headquartered company with no published pricing and a smaller public profile than the venture-funded US and Singapore competitors here.
What stands out
EU-founded, Leipzig
Knowledge-graph architecture
Low-code for subject-matter experts
GermanyQuote-only, no published tiers; a free community edition sandbox (eccenca.my) lets you try it before a demo
Open-source catalog, lineage and governance platform with a commercial managed tier from Collate
Ranked #2 of 11 in Best Data Catalog Software in 2026.
Free tierOpen sourceSelf-hostablePublished pricingNorth America
The project page lists 130+ connectors and describes a four-component architecture meant to be easy to deploy and upgrade. Collate adds SSO and PII classification from Premium, with single-tenant or hybrid SaaS, while the free tier is multi-tenant SaaS.
Exit is the strongest on this list, since the metadata sits in a system you run under an open licence. The cost is staff time: someone must own installation, upgrades and backups.
What stands out
Open source
Self-hostable
14,000+ open-source members
Where it costs you
Self-hosting means your own staff patch, upgrade and monitor the stack.
The Collate free tier stops at 500 data assets, which a mid-sized warehouse can pass in a single schema.
Right for
Engineering-led teams that want open code and the option of a hosted tier
Wrong for
Business-led governance programmes with no engineers to run it
United StatesFree self-hosted open source; Collate's managed tier has a free plan (5 users, 500 assets), Premium and Enterprise are quote-only
All-in-one catalog, governance and AI data analyst, acquired by Atlassian in December 2025
Ranked #3 of 11 in Best Data Catalog Software in 2026.
Pricing on requestElsewhere
Per the pricing page, Enterprise adds SIEM logging, unlimited integrations, disaster recovery support and professional services. Core already includes API access, automations and workspace analytics. No amounts appear on the page.
Ownership is the buying risk: contracts, roadmap and data handling now sit with a larger company, and the page does not describe export. Ownership by a large company is why it ranks below the smaller vendors on independence.
What stands out
Catalog + lineage + observability in one
24/7 AI data analyst
Acquired by Atlassian, Dec 2025
Where it costs you
The vendor is now owned by Atlassian, and the pages read do not say how the product will evolve.
Policies, PII scanning and single-tenant deployment need Premium or above.
Right for
Data teams already inside the Atlassian ecosystem wanting lineage and catalogue
Wrong for
Buyers wanting an independent vendor with a stated country
Not stated by the vendorQuote-only across Core, Premium and Enterprise tiers
Singapore-based active metadata platform selling itself as AI agents' missing context layer
Ranked #4 of 11 in Best Data Catalog Software in 2026.
Pricing on requestAsia-Pacific
Atlan reverse-engineers a live metadata graph out of a company's warehouses, BI tools and pipelines, then exposes it through open APIs so both people and AI agents query the same governed context instead of guessing at a table name. It raised a $105M Series C in May 2024 at a $750M valuation led by Singapore's GIC, so it is well-capitalised and not going away.
The catch is the same as everywhere else in this category: no published price, and the AI-agent framing is where the roadmap is pointed, which means a buyer evaluating it purely as a classic catalog is not seeing the product at its strongest.
The data catalog pioneer, now selling governed context to AI agents as much as to analysts
Ranked #5 of 11 in Best Data Catalog Software in 2026.
Pricing on requestNorth America
Alation's site is organised as an operating system with modules rather than as one catalogue: catalogue, lineage, quality, curation, Critical Data Manager, data products, marketplace and analytics.
More modules means more decisions about what to buy first. The pages read give no limits, tiers or export terms, so those must be requested in writing before a pilot, along with the list of modules the first-year price covers.
What stands out
Catalog pioneer
AI governance layer
Quote-only
Where it costs you
The pricing page names no plan, price or trial, so comparison starts only after a sales call.
The headquarters country is not stated on the pages read.
Right for
Large organisations that expect a sales-led rollout across many data sources
Lyon-founded catalog built for business users first, technical governance second
Ranked #6 of 11 in Best Data Catalog Software in 2026.
Pricing on requestEurope
The French company was founded in Lyon in 2015, added a Paris office in 2019 and, per its own history page, acquired Yooi in 2025. Starter includes one production instance and SSO.
Because every tier is quoted and editors have a floor, the real bill depends on how many stewards you name. The pricing page does not describe export or any self-hosted option.
What stands out
EU-founded, Lyon, 2015
Business-user-first design
Catalog + governance + AI-initiative tracking
Where it costs you
Tiers carry editor floors of five, fifteen and fifty, so small teams start at a quote.
Starter is capped at two workspaces and five core connectors.
Right for
Mid-sized firms wanting a business-friendly catalogue with a governance module
Wrong for
Teams needing public prices or a self-serve trial
FranceQuote-only across Starter, Professional and Enterprise catalog tiers, plus a separate Portfolio product line
Quest-owned catalog and governance suite that also covers unstructured data and AI-model certification
Ranked #7 of 11 in Best Data Catalog Software in 2026.
Pricing on requestNorth America
erwin's data-modelling roots go back to 1984, and Quest Software bought the company in October 2021 and still sells it as Quest erwin. The genuine differentiator against the pure-play catalogs on this list: it explicitly governs unstructured data and can certify an AI model's lineage, not just tag warehouse tables.
It is a smaller, less AI-agent-focused brand than Alation or Atlan now, and being one product inside a much larger Quest portfolio means the catalog is not the company's full attention the way it is for an independent vendor.
The Belgian-founded catalog and governance incumbent, now dual-headquartered in New York and Brussels
Ranked #8 of 11 in Best Data Catalog Software in 2026.
Pricing on requestElsewhere
Collibra's homepage frames the product as context and control for any model, data and agent, and groups it by industry. The pages read give no tier limits, plan names or export terms.
Independence is the weak point for a small buyer: the only route in is a demo request, and the cost of the first ninety days stays unknown until a sales team scopes it.
What stands out
Belgian-founded, 2008
Dual New York/Brussels HQ
AI Command Center / Context Engine
Where it costs you
No price, plan names or trial are shown on the pages read.
The suite is aimed at large programmes, so smaller teams carry setup they may not need.
Right for
Large enterprises running governance across many business units and regions
Wrong for
Small teams needing a quick and cheap start
Belgium / United States (dual HQ)Quote-only; no published tiers
Austin-founded knowledge-graph catalog, now part of ServiceNow
Ranked #9 of 11 in Best Data Catalog Software in 2026.
Pricing on requestNorth America
data.world built its catalog and governance product on a knowledge-graph architecture rather than a flat metadata table, aimed at connecting business meaning to technical assets.
Its own site now states plainly that "data.world is now part of ServiceNow," so it is no longer evaluated as an independent company; expect its roadmap, pricing and packaging to fold into ServiceNow's much larger enterprise-platform sales motion rather than continue as a standalone product with its own pricing page.
Microsoft's business-facing data catalog, layered over the Purview Data Map metadata graph
Ranked #10 of 11 in Best Data Catalog Software in 2026.
Free tierPublished pricingNorth America
Unified Catalog is the layer of Microsoft Purview where a data product, a business glossary term and the technical tables behind it get connected, sitting on top of Purview's Data Map scanning engine. It began as Azure Purview, whose original portal and pricing are now labelled "classic", and it replaced Azure Data Catalog, a separate service retired in May 2024, which makes it easy to find outdated documentation.
Pricing is genuinely metered and published rather than quote-only, and a free version handles basic Azure and Fabric cataloguing up to 1,000 annotated assets, but it only makes sense as a purchase for an organisation already committed to Azure and Microsoft 365; it is not a portable, engine-agnostic catalog.
What stands out
Consumption-based, published rates
Governs data, AI models and M365 content together
Formerly Azure Purview; replaced Azure Data Catalog
United StatesFree tier covers browse and search of Azure and Fabric sources up to 1,000 annotated assets; enterprise is consumption-based, billed per governed asset per day plus Data Governance Processing Units for data-quality rules, with published pay-as-you-go rates
Google's Gemini-powered semantic catalog, renamed from Dataplex Universal Catalog in April 2026
Ranked #11 of 11 in Best Data Catalog Software in 2026.
Published pricingNorth America
Google's catalog has a tangled lineage: Dataplex Universal Catalog, which replaced the older Data Catalog service, was renamed Knowledge Catalog on 10 April 2026, with the API, client library, CLI and IAM names left as Dataplex underneath. The current pitch is an active context graph that grounds Gemini and other AI agents in a company's real metadata, across both structured tables and unstructured files, rather than a passive list of column names.
The naming churn is the real risk here: documentation, tutorials and job postings referencing "Dataplex" are already out of step with the product's current name, and while per-unit rates are published, the bill depends on processing volume that is hard to estimate before a pilot.
What stands out
Renamed from Dataplex, April 2026
Gemini-powered semantic context
Structured + unstructured coverage
United StatesPay-as-you-go with published rates per DCU-hour of processing and per GiB of metadata storage; 100 DCU-hours of standard processing a month are free
Data catalog software inventories an organization's data assets, tables, dashboards, files and models, and records what each one means, who owns it and where it came from. Data governance software layers policy on top of that inventory: access rules, stewardship workflows and an audit trail, and increasingly, the governed context an AI agent reads before it answers a question about the business. A feature table will not settle this choice. These steps will.
01
Start with pricing transparency
Only 4 of the 11 print a price; 7 quote per organisation. Comparable without a sales call: eccenca Corporate Memory, OpenMetadata, Microsoft Purview Unified Catalog, Google Cloud Knowledge Catalog.
02
Look at where the vendor sits
Look at the contracting entity: these 11 vendors sit in 6 countries, split by region as North America 6, Europe 2, Elsewhere 2, Asia-Pacific 1. The servers can be anywhere; the legal obligations follow the company.
03
Decide if open source matters
1 of these are open source: OpenMetadata. Hosting one yourself trades a subscription for maintenance.
Inside one cloud, or above all of them
Microsoft Purview Unified Catalog and Google Cloud Knowledge Catalog are metered services within their own clouds, and Purview only makes sense for an organisation committed to Azure and Microsoft 365.
The independent vendors, Atlan, Alation, Collibra, DataGalaxy and OpenMetadata among them, are not tied to one provider. What matters is whether your data estate will plausibly span several clouds during the catalog's lifetime.
List every warehouse, lake and BI tool, then ask each vendor which have a native connector.
For Purview, estimate the governed asset count, since billing is per asset per day plus processing units for data-quality rules.
Search Google's documentation under both Dataplex and Knowledge Catalog, because the API names never changed.
Who is meant to open it: stewards, engineers or agents
DataGalaxy designs for business analysts and stewards first, and eccenca aims its low-code tooling at subject-matter experts. OpenMetadata, with lineage, quality and observability in one graph, appeals to engineering teams.
Atlan, Alation and Collibra now pitch the catalog as governed context for AI agents. A catalog filled by engineers and ignored by the business documents tables without ever settling what a term means.
Have a business analyst, not an engineer, look up a metric definition during evaluation.
Ask how an AI agent reads the glossary: through open APIs, as Atlan describes, or something proprietary.
Decide who approves a glossary term and confirm the workflow for it exists.
Several catalogs here now answer to a larger parent
Secoda was acquired by Atlassian in December 2025, data.world is now part of ServiceNow, and erwin Data Intelligence has belonged to Quest Software since 2021. In each case roadmap and packaging follow the parent's priorities.
Google renamed Dataplex Universal Catalog to Knowledge Catalog in April 2026, and both Google and Microsoft have replaced an older catalog service with the current one. A catalog is a multi-year commitment, so the parent's intentions belong in the evaluation.
Ask Secoda and data.world whether the product will still be sold standalone in two years.
Verify that the documentation you rely on uses the current product name.
Ask how much of the parent company's roadmap is dedicated to the catalog itself.
Flat inventory or knowledge graph, hosted or your own
eccenca Corporate Memory and data.world model metadata as a knowledge graph, which captures business rules and relationships that a list of tables cannot, in exchange for more modelling effort.
erwin also governs unstructured data and certifies AI-model lineage. OpenMetadata is the one option whose source you can read and self-host; Collate's free managed plan stops at 5 users and 500 assets.
Ask whether lineage and glossary terms leave through an open API or only as flat files.
Count assets against Collate's 500-asset free plan before assuming it covers a pilot.
If documents and files must be governed, confirm unstructured coverage, which erwin and Google's catalog both state.
What goes wrong most often when buying data catalog software
Running the catalog as an engineering project and expecting business users to fill in the glossary afterwards.
Picking Microsoft Purview for a multi-cloud estate when it is built for Azure and Microsoft 365.
Evaluating Secoda or data.world as independent products after their acquisitions by Atlassian and ServiceNow.
Self-hosting OpenMetadata to avoid licence fees without budgeting for the ongoing operational work.
07
Frequently asked questions
8 answers
What is the best data catalog software in 2026?
Our ranking of 11 puts eccenca Corporate Memory first. eccenca's Corporate Memory models an organisation's data and business rules as a semantic knowledge graph rather than a flat catalog table, with a stated goal of making the result explainable enough for AI systems built on top of it to be trusted.
Low-code tooling is aimed at subject-matter experts rather than only data engineers, which is a genuine differentiator from the more developer-first tools on this list. It is a smaller, German-only-headquartered company with no published pricing and a smaller public profile than the venture-funded US and Singapore competitors here.
Which data catalog tools publish their pricing?
4 out of 11 state a price model on their own site:
eccenca Corporate Memory: Quote-only, no published tiers; a free community edition sandbox (eccenca.my) lets you try it before a demo.
OpenMetadata: Free self-hosted open source; Collate's managed tier has a free plan (5 users, 500 assets), Premium and Enterprise are quote-only.
Microsoft Purview Unified Catalog: Free tier covers browse and search of Azure and Fabric sources up to 1,000 annotated assets; enterprise is consumption-based, billed per governed asset per day plus Data Governance Processing Units for data-quality rules, with published pay-as-you-go rates.
Google Cloud Knowledge Catalog: Pay-as-you-go with published rates per DCU-hour of processing and per GiB of metadata storage; 100 DCU-hours of standard processing a month are free.
The remaining 7 sell by quote.
Is there a free data catalog tool?
OpenMetadata, Microsoft Purview Unified Catalog offer a free tier or a free self-hosted edition.
Where are these data catalog software vendors established?
In 6 countries across 4 regions: North America 6, Europe 2, Elsewhere 2, Asia-Pacific 1.
eccenca Corporate Memory: Germany.
OpenMetadata: United States.
Secoda: Not stated by the vendor.
Atlan: Singapore.
Alation: United States.
DataGalaxy: France.
erwin Data Intelligence: United States.
Collibra: Belgium / United States (dual HQ).
data.world: United States.
Microsoft Purview Unified Catalog: United States.
Google Cloud Knowledge Catalog: United States.
Which data catalog tools are open source?
OpenMetadata.
Which data catalog tools can you host yourself?
OpenMetadata. The remaining 10 run only on the vendor's servers.
What should you use instead of eccenca Corporate Memory?
Second is OpenMetadata, which suits Engineering-led teams that want open code and the option of a hosted tier. Third is Secoda, which suits Data teams already inside the Atlassian ecosystem wanting lineage and catalogue.
Who should not buy eccenca Corporate Memory?
eccenca's Corporate Memory models an organisation's data and business rules as a semantic knowledge graph rather than a flat catalog table, with a stated goal of making the result explainable enough for AI systems built on top of it to be trusted.
Low-code tooling is aimed at subject-matter experts rather than only data engineers, which is a genuine differentiator from the more developer-first tools on this list. It is a smaller, German-only-headquartered company with no published pricing and a smaller public profile than the venture-funded US and Singapore competitors here.
Missing from this data catalog software ranking? Contact us with a short description of the product and its users. Adding it is an editorial choice, and software advice explains what a listing does and does not involve.
