Why DataMind

Solutions

Company

DataMind Universe

Free trial

The whole OS. Two months. Free.

Every app, on your own servers, with your own data.

  • Your data stays on your servers On your servers
  • Installed in about 20 minutes 20-minute install
  • No card, no feature limits No card

One quick check, then it sends.

Your request goes to our sales team. How we handle it (opens in a new tab)

Ask DataMind

DataMind:

Hello, Winner.

Tell me who you are and what slows you down. I will show you how DataMind OS fixes it.

This chat is not stored. Answers run on zero-data-retention model endpoints. Privacy

DataMind OS vs Looker

Everything Looker does. In SQL, on your servers.

Model in plain SQL instead of LookML: DataPrism, the BI of DataMind OS, does the work teams use Looker for, and Report Builder AI rebuilds each dashboard from its screenshot in minutes, on your own servers.

01Verdict

The short answer.

DataMind OS covers everything teams use Looker for, and adds much more on top: the lake, the warehouse and workflows under the semantic layer, Report Builder AI, Slides and more than 20 AI agents, in one OS on your own servers at one company price, with every report builder and viewer free.

Looker today. Google Cloud's business intelligence and semantic modeling service (LookML), mainly hosted by Google, that queries a SQL database the customer runs separately.

  • On your servers.

    Installed on your own Linux servers in about 20 minutes, in any cloud or data center.

  • One OS, one login.

    Connectors, Lake, Warehouse, DataPrism BI, Slides and AI Minds share one set of access rules in the database.

  • Free builders and viewers.

    Report and dashboard builders, Mind builders, viewers and subscribers are free for an unlimited number of users.

02Side by side

DataMind OS and Looker, row by row.

14 topics. Every Looker fact cites a numbered source, checked October 2026.

DataMind OS compared with Looker, checked October 2026
Topic DataMind OS Looker
Where it runs DataMind OS installs on your own Linux servers, physical or virtual, in any cloud account or data center, in about 20 minutes through the browser Installer. The Lake, the Warehouse, models and reports live on your servers, and a named release arrives every month, installed with one click.4 Looker (Google Cloud core) is hosted by Google in Google Cloud, in 43 regions, and Google states it is not available customer-hosted or multicloud. Looker (original) can still be customer-hosted, without Gemini in Looker or the Looker-managed MCP server, and the customer runs updates, scaling and uptime.1, 2, 3
Price One published company plan, on the pricing page. Every app, SSO and two-factor and Kartlos tokens come with it, plus one Privileged seat for integrations, the data warehouse and access permissions.6 A platform fee for the Standard, Enterprise or Embed edition plus user licenses, quoted by Google sales on one to three year terms.5
Users and licenses Report and dashboard builders, Mind builders, Oda users, viewers and subscribers are free, for an unlimited number of users. With 1,000 report developers and 10,000 viewers, those 11,000 seats cost nothing.6 Every person holds a Developer, Standard or Viewer license. Each edition includes 10 Standard and 2 Developer users, the Standard edition stops at 50 users, and non-production instances are bought separately.5, 7, 1
Data foundation Connect Anything, Lake, Warehouse, transformations and Workflows ship in the same OS. Your data lands in ClickHouse on your own servers, columnar and ZSTD-compressed, and every chart, export and subscription queries it live, with no extracts. Looker is a BI and semantic layer over a SQL database you run, such as BigQuery, Snowflake, Redshift or Databricks. It queries that database in place, without extracts.8, 9
Modeling language SQL, the language your analysts and AI already write. Measures and calculated columns are ClickHouse SQL with window functions and 130 recipes, AI writes one from a sentence, and each measure carries up to 50 tests on live data, so a Model AI change that breaks a number is refused.13 LookML, Looker's own modeling language, written by Developer users in Git-backed project files; Looker generates the SQL. Continuous Integration (GA) validates SQL and content before changes ship.10, 11, 12
AI that builds reports Report Builder AI builds report pages from a description or a screenshot in minutes: one plan card, one approval, 57 report tools and a Verifier that checks the pages. It asks Model AI for missing measures and hands bespoke tiles to Custom Visual AI.15 The Visualization Assistant (GA) formats charts from a prompt; the Expression Assistant and Insight Assistant are in preview, and Dashboard Agents (preview) answer questions inside a dashboard.14, 12
Custom visuals Custom interactive visuals you create with AI on the VDK (Visual Development Kit): Custom Visual AI writes a tile from a sentence or a picture, the custom visual Studio gives live preview and 12 libraries such as d3, ECharts and three.js, and the visuals cross-filter, act as slicers and export to PPTX and PDF.17 Custom visualizations are JavaScript files added through the LookML manifest, the Looker Marketplace or the admin panel; Google describes them as a community-supported effort.16
AI analysts AI Minds answer by writing and running SQL on governed tables, every step a card with its SQL and result. Execute as User runs each query under the asking person's own rules, and Swarms, Scheduled Tasks and Benchmarks on your own questions come in the same plan. Conversational Analytics (GA) answers questions on Explores and dashboards under Looker permissions, and Code Interpreter runs Python for advanced analysis. Looker BI Agents and Agentic Workflows were announced at Next '26, several in preview.14, 12
AI and your data Zero data retention: model calls go through DataMind's AI gateway to zero-data-retention endpoints, so nothing a Mind sends is stored by DataMind or the model provider. Columns marked as personal data reach the model as tokens, a Region setting routes AI chat to EU endpoints, and every AI request is logged in your own database. Google does not train on prompts or company data. Conversation data at rest stays in the instance's region while data in transit and ML run on a global service; conversations are kept until a user deletes them, and Gemini is not supported on instances with CMEK or VPC Service Controls.14, 18
AI tokens 85M Kartlos Premium and 850M Kartlos Standard tokens a month come with the plan, about 1B and 10B a year.6 Conversational Analytics draws on a monthly data-token quota per edition, from 60M input and 1.2M output tokens on Standard to 1.2B and 24M on Embed. Today's use is promotional; overage fees start after at least 90 days notice.5
Connectors 22 built-in source types for databases, cloud warehouses, files, queues and webhooks, BigQuery, Snowflake, Redshift, Athena and Azure SQL among them. Connect Any API has AI build a connector for any REST, OData or GraphQL API, and the custom connector agent writes and tests a Python connector for anything else. Looker reads SQL databases in place: 39 products at its Supported level, among them BigQuery, Snowflake, Redshift, Databricks, SQL Server, Oracle, SAP HANA and ClickHouse. Self-service Explores (GA) blend uploaded CSV and Excel files with modeled data.9, 12
Row and column security Object, column and row rules live in the database, so they hold in DataPrism, Minds, the SQL Editor and in DataGrip, Power BI or Excel. A semantic model limits rows per viewer by group or by a SQL rule, and View as shows a report exactly as a colleague sees it.20 Row restrictions are set per Explore in LookML with access_filter and user attributes, and roles, permission sets and model sets control what each user can do.19, 14
Delivery and exports Email subscriptions deliver PDF, PowerPoint or Excel up to 24 times a day, each copy under its recipient's row rules. Every visual streams its whole result natively, even a billion rows as CSV or TSV, and every chart is an API in CSV, JSON or Parquet.23, 24 Schedules deliver to email, webhook, Amazon S3 or SFTP, with dashboards as PDF, visualization or CSV zip. Queries return 5,000 rows by default, and downloading all results needs the download_without_limit permission and a streaming result.21, 22
Time to value About 20 minutes from a bare Linux server to the whole OS, with a demo semantic model and report seeded on day one. The two-month free trial runs the whole OS on your own servers, with no card and no feature limits.4 A trial instance runs 90 days, and production needs a new instance plus an import. Models are written in LookML by Developer users, and Google publishes no time-to-production figure.1, 7

Checked . Sources at the foot of the page.

03Further

Where DataMind OS goes further.

All of it in the same plan, on the same servers, under the same rules.

  1. SQL instead of LookML

    Measures and calculated columns are ClickHouse SQL, and SQL is the better language for both: 130 recipes in the Measure builder, AI that writes one from a sentence and up to 50 tests per measure on live data.

  2. Report Builder AI

    Builds report pages from a description or a screenshot in minutes, with one plan card, one approval and a Verifier that checks the pages.

  3. The data foundation included

    Connect Anything, Lake, Warehouse, transformations and Workflows ship in the same OS, on ClickHouse on your own servers.

  4. Custom Visual AI on the VDK

    Custom Visual AI writes custom interactive visuals on the VDK, the Visual Development Kit, that cross-filter, act as slicers and export to PDF and PPTX.

  5. Native streaming download

    Every visual streams its whole result straight from ClickHouse, even a billion rows as CSV or TSV, and every chart is also an API in CSV, JSON or Parquet.

  6. Free builders and viewers

    No Developer, Standard or Viewer licenses: 1,000 report developers and 10,000 viewers cost nothing.

  7. Slides

    Slides AI builds and edits decks whose charts, tables and text bind to live SQL, and decks export to PowerPoint with native, editable charts.

04Switch

Switch from Looker.

Your warehouse stays where it is and connects on day one, each Looker dashboard comes over from its screenshot in minutes, and Looker keeps running on the same data until the last page moves.

  1. Day one

    Install DataMind OS on your own Linux server in about 20 minutes; a demo semantic model and report are ready before any source connects. The warehouse Looker reads connects as a Lake source: BigQuery, Snowflake, Redshift, Athena, Azure SQL, Postgres, SQL Server, MySQL, Oracle and ClickHouse are among the 22 built-in types, SQL database and Snowflake sources also take your own SQL query as a custom table, and the Lake Builder Mind imports up to 1,000 tables at once under your approval.

  2. The model

    Create a semantic model on those tables. Model AI adds the BI tables, sets the fact table and the joins, and writes each measure in SQL from a plain-language description of the LookML measure, each change on a card with an undo. A measure test compares the new measure with your own source-of-truth SQL before anyone relies on it, and /star-schema in the Warehouse Builder builds a governed star schema checked against totals you trust when you want one.

  3. The dashboards

    Put a screenshot of each Looker dashboard into New report with AI. Report Builder AI marks every element on the plan card, you say what each mark becomes, and it rebuilds the page on your model in minutes; Custom Visual AI rebuilds a bespoke visual from its picture. From then on a report travels as a .dataprism file and a model as a .semanticModel file.

  4. Rules and delivery

    The row restrictions each access_filter applied are rebuilt as the model's row-level security, by group or by a SQL rule on the viewer's email, and View as checks each colleague's view. Each Looker schedule becomes an Email subscription in PDF, PowerPoint or Excel, each copy under its recipient's row rules.

  5. In parallel

    Looker keeps querying your warehouse untouched while pages move, because DataMind OS connects to the same sources. A Consulting and delivery partner can run the whole program with your team.

Q and A

Questions, answered.

What teams ask before they move from Looker.

Is there a self-hosted alternative to Looker?

Yes. DataMind OS installs on your own Linux servers, in any cloud account or data center, in about 20 minutes, and brings the Lake, the Warehouse, DataPrism BI, Slides and AI Minds in one OS. Your data, models and reports stay on your servers, and every chart queries your own ClickHouse live.

How is DataMind OS priced compared with Looker?

DataMind OS is one company plan, published on the pricing page, with one Privileged seat included. Report builders, Mind builders, Oda users, viewers and subscribers are free at any scale, and 85M Kartlos Premium and 850M Kartlos Standard tokens a month are included. Google quotes Looker per edition and per user through its sales team.

Do we need LookML or another modeling language?

No. Measures and calculated columns are SQL, the language analysts and AI already write. The Measure builder offers 130 recipes, AI writes a measure or a column from a sentence, and each measure carries up to 50 tests on live data, so the numbers stay right through every change.

Can we move our Looker dashboards to DataMind OS?

Yes. Each page comes over from its screenshot: New report with AI or Report Builder rebuilds it on your semantic model in minutes, Model AI writes the measures in SQL, and Custom Visual AI rebuilds bespoke visuals from their pictures. Looker keeps running on the same warehouse until the last page moves.

Does DataMind OS work with BigQuery?

Yes. BigQuery connects as a built-in Lake source, beside Snowflake, Redshift, Athena, Azure SQL and the rest of the 22 source types, and BigQuery is also a Writeback destination for modeled data.

How does the AI treat our data?

DataMind is zero data retention. Model calls go through DataMind's AI gateway to zero-data-retention endpoints, so nothing a Mind sends is stored by DataMind or the model provider. Columns marked as personal data reach the model as tokens while your people see the real values, and every AI request is logged in your own database.

Can we try DataMind OS on our own data while Looker keeps running?

Yes. The two-month free trial runs the whole OS on your own servers, with no card and no feature limits. Your warehouse connects on day one, Looker keeps running on the same data, and your data stays on your servers either way.

05Sources

Where the facts come from.

Checked . Product names are trademarks of their owners.

  1. Looker (Google Cloud core) overview (updated 2026)
  2. Looker (Google Cloud core) locations (updated 2026-10-26)
  3. Choosing a hosting option for a Looker (original) instance
  4. DataMind docs: Install DataMind OS
  5. Looker pricing
  6. DataMind OS pricing
  7. Looker (Google Cloud core) editions and user types (updated 2026-10-26)
  8. Looker product page
  9. Looker dialect support (Looker 26.20, updated 2026-10-26)
  10. Introduction to LookML
  11. LookML terms and concepts
  12. Welcome to the agentic BI era with Looker (April 23, 2026)
  13. DataMind docs: Measures
  14. Gemini in Looker overview (updated 2026-10-22)
  15. DataMind docs: Report Builder AI
  16. Admin settings: Visualizations (updated 2026-10-26)
  17. DataMind docs: Custom visuals
  18. Understanding Looker Conversational Analytics security (July 8, 2025)
  19. LookML access_filter
  20. DataMind docs: Connection settings
  21. Using the Looker Scheduler to deliver content
  22. What are all the row limits in Looker? (updated 2026-10-22)
  23. DataMind docs: Export and the chart API
  24. DataMind docs: Subscriptions
  25. Create a Looker (Google Cloud core) instance (updated 2026-10-26)
  26. Self-service migration from Looker (original) to Looker (Google Cloud core)
  27. Toucan, Looker Pricing 2026 (third party, updated June 2026)
  28. Looker a Leader in the 2026 Gartner Magic Quadrant (July 2, 2026)
  29. Looker Studio is Data Studio (April 2026)

Try it beside Looker.

Two months, the whole OS, on your servers. Free.