by Google Cloud

Looker review, pricing and verdict

Google Cloud's BI platform built on a LookML semantic layer, with Gemini conversational analytics that is unmetered for now and metered from 2027.

Best for
Data engineer
Not for
Business user who wants to self-serve without a developer

Should you use Looker?

Choose Looker if you run Google Cloud or BigQuery and want one governed metrics layer for people and AI agents. It replaces scattered dashboard definitions. Skip it if you need a published price or a tool business users can set up alone, because LookML needs developers.

A LookML semantic layer that both dashboards and Gemini agents read.

Also for

  • BI lead on Google Cloud
  • Data engineers who maintain a governed metrics layer
Show 3 more
  • BI leads at Google Cloud and BigQuery companies
  • Product managers embedding analytics in a SaaS product
  • Analytics teams building AI agents on governed data

Main trade-off

No published platform price, and the best AI features depend on how the instance is configured.

Looker pricing

PlanMonthly priceWhat it includes
Looker StandardCustomFor teams with fewer than 50 users, Includes the base bundle: one production instance, 10 Standard Users, 2 Developer Users, Up to 1,000 query-based API calls a month, Up to 1,000 administrative API calls a month, Annual terms, call sales
Looker EnterpriseCustomEnhanced security features for internal BI, Same base bundle as Standard, Up to 100,000 query-based API calls a month, Up to 10,000 administrative API calls a month, Quoted by sales
Looker EmbedCustomFor external analytics and custom applications, Same base bundle as Standard, Highest API limits of the three editions, Up to 100,000 administrative API calls a month, Price on request from Google Cloud sales

Platform editions (Standard, Enterprise, Embed) are annual and quoted by Google Cloud sales, in one, two or three year terms. Conversational Analytics is unmetered until 31 January 2027, then billed beyond the included pool at $3.00 per 1M input tokens and $20.00 per 1M output tokens.

Before you commit

  • Conversational Analytics beyond the included pool is metered from 1 February 2027
  • LookML needs developer time, and Developer Users are a separate licence type

What Looker does

Looker is Google Cloud's business intelligence platform. Analysts write business logic once in LookML, and dashboards, embedded apps and Gemini-powered Conversational Analytics all read it. It runs as managed Looker (Google Cloud core) in Standard, Enterprise and Embed editions, each including 10 Standard Users and 2 Developer Users.

  • LookML semantic layer

    Business logic is defined once in LookML and reused by dashboards, the API, embedded reports and AI agents.

  • Conversational Analytics

    Gemini turns plain-language questions into governed Looker queries, in Explores and Dashboards or through the Conversational Analytics API.

  • Gemini assistance

    Google's editions matrix lists Conversational Analytics, Code Interpreter, LookML writing assistance and a Visualization Assistant under Gemini in Looker.

  • Looker Embedded

    The Embed edition is the one meant for deploying external analytics and custom applications at scale.

Show all 7 features
  • Looker APIs and MCP

    Extensible APIs and MCP integration let other tools and agents query the same governed metrics.

  • BigQuery integration

    Looker is positioned as the semantic modelling layer over BigQuery data.

  • Three user licences

    Developer, Standard and Viewer users get different interfaces, from LookML development down to view-only dashboards.

Works with

  • BigQuery
  • Gemini
  • Google Sheets
  • Power BI connector
  • Tableau connector

Capabilities

  • Business intelligence
  • Semantic layer
  • Conversational analytics
  • Embedded analytics
  • Dashboards
  • Lookml

Strengths and limits

Strengths

  • One LookML definition feeds dashboards, the API and Gemini answers, so the number an AI agent quotes matches the number on a dashboard, per Google Cloud's Looker page read 22 November 2026.
  • Each edition includes 10 Standard Users and 2 Developer Users, so a small team can start without buying seats one by one, per Google Cloud's pricing page of the same date.
  • Conversational Analytics carries no allocation limit or extra billing, within fair use, until 31 January 2027, so a buyer can measure real usage first.
  • Query-based API allowances rise from 1,000 a month on Standard to 100,000 on Enterprise and 500,000 on Embed, a clear step for customer-facing products.

Limits

  • No platform price is published: Standard, Enterprise and Embed all read "Call sales" with annual terms, so a buyer cannot budget from the page.
  • Google's editions documentation, read 22 November 2026, says Gemini assistance is not supported on Looker (Google Cloud core) instances that use CMEK or VPC Service Controls, which can rule it out for strict security setups.
  • From February 2027 Conversational Analytics use above the included pool is billed per token, so heavy use turns into a variable monthly cost for the buyer.
  • The Standard edition is meant for teams with fewer than 50 users, so a growing company will need Enterprise and a new quote.

How Hokai rates it

  • Ease of use6.5 / 10
  • Features8.5 / 10

Looker alternatives

ToolPrice fromFree planRuns onRating
Looker This pageQuoteNoWeb3.8
TableauPick Looker if developers should define governed metrics once for people and agents; pick Tableau if analysts want visual-first authoring and a free Desktop edition.$15/moYesWeb3.8
Microsoft Power BIPick Looker if you are on Google Cloud and can take a sales quote; pick Power BI if you want published per-user prices and a free authoring app.$14/moYesWeb · iOS · Android · Windows3.8
ThoughtSpotPick Looker for a LookML semantic layer and embedded analytics; pick ThoughtSpot if business users should pull their own answers from live cloud data.$25/moYesWeb--

Looker facts

Works offline
No
Minimum skill level
Advanced
Price band
Enterprise pricing

Data and privacy

Proprietary

Questions about Looker

How much does Looker cost?

Google Cloud does not publish a platform price. Standard, Enterprise and Embed each read call sales, with one, two or three year annual terms, and each includes 10 Standard Users and 2 Developer Users. Conversational Analytics above the included pool is billed from 1 February 2027, at the token rates listed under price notes.

Is Looker free to use?

No, there is no free plan. Google offers a free trial, with one active trial per edition for each Google Cloud project and region. Conversational Analytics has promotional use with no allocation limits or extra billing, within fair use, through 31 January 2027.

What do Looker's Standard, Enterprise and Embed editions include?

Each includes one production instance, 10 Standard Users and 2 Developer Users. They differ in API limits and purpose: Standard is meant for teams under 50 users, Enterprise adds security features for internal BI, and Embed targets customer-facing analytics. Query-based API calls are capped at 1,000 a month on the smallest edition and rise sharply on the other two.

What is the difference between Looker Developer, Standard and Viewer users?

A Developer User can reach every interface, including Administration, LookML models in Development Mode and the API. A Standard User gets folders, dashboards, Looks, Explore, SQL Runner and scheduling. A Viewer User can open folders, boards, dashboards and Looks, filter and download data.

Does Gemini in Looker work on every instance?

No. Google's editions documentation says Gemini assistance is unavailable on instances configured with CMEK (customer-managed encryption keys) or VPC Service Controls. Publishing data agents to Gemini Enterprise also is not supported where an instance uses private or hybrid network connections.

About Looker

Looker is the business intelligence platform from Google Cloud. Its core idea is a semantic layer written in a modelling language called LookML: analysts define business logic once, and dashboards, embedded reports, the API and Gemini-powered Conversational Analytics all read the same definitions. Google's own page calls it "the agentic BI platform built on a trusted semantic layer" and says it connects to MCP-based tools.

Read the full background

It is sold as Looker (Google Cloud core), a managed instance in three editions: Standard for small teams with fewer than 50 users, Enterprise for internal BI with extra security features, and Embed for customer-facing analytics. Each edition starts from the same base bundle of a production instance plus 10 Standard and 2 Developer seats. Looker (original) also exists as a customer-hosted version.

It suits data teams who already run BigQuery or another warehouse and want one governed set of metrics, and software companies that embed analytics in their own product. The people who write LookML are developers, so a Looker rollout is closer to a data engineering project than a drag-and-drop dashboard purchase. Viewers, by contrast, only open dashboards and filter data, so the cost of a wide rollout sits mostly in the few people who model the data.

It is a poor first choice for a team that wants a price on a web page. Every platform edition is annual and call-sales, and conversational analytics moves to metered billing in February 2027. A company inside Microsoft 365 may prefer Microsoft Power BI, a visual-first team may prefer Tableau, and a business-user search tool may fit ThoughtSpot better (see the data science and ML platforms list).

The call: Looker fits a Google Cloud shop that needs one semantic layer for people and AI agents, or a software company embedding analytics for customers. Budget for developers and a sales quote first, and read the best AI data analytics tools guide before you shortlist it.

Sources and guides

More AI Tools on Hokai

  • DeepnoteShared cloud notebook with Python and SQL blocks, an AI agent and an MCP server; the Team plan includes $39 of AI credits per editor each month.
  • HexCollaborative data workspace for analysts: code notebooks, an agent and published data apps; the vendor blog of 27 October 2026 cites 2,500 customers.
  • AkkioAkkio is an AI platform for media agencies, with agents for audience segments, media-plan forecasts, activation and measurement. One custom-priced enterprise plan; no free tier.
  • EncordData infrastructure for Physical AI and multimodal teams: annotation, curation, alignment and managed data, used by 300+ AI teams, per the vendor.
  • InfuseBIGoverned BI workspace where six AI agents model data, explain changes, forecast and draft dashboards, embeddable in SaaS apps, with a 15-day trial.
  • Superb AISeoul-based vision AI vendor selling data labeling, curation, model training and video analytics to industrial and mobility teams, all quote-only.