All roundups

Alternatives

Best Looker alternatives

Looker is two things: a semantic model written in LookML, and a BI tool sitting on it. The model is why people buy it and why they leave. Somebody writes it, then keeps writing it, and Google publishes no price for any of it, so a quote and a 1, 2 or 3 year term come before the first dashboard does. Here are nine alternatives, grouped by the job they suit.

How to read this

Flitch publishes this page and is one of the tools on it. Prices are the vendors’ own published figures, and every entry says what the tool costs you as well as what it is good at.

Prices checked

At a glance

ToolBest forEntry price
Power BICompanies already running Microsoft 365, Azure or Fabric.Free account, then $14/user/month paid yearly
TableauAnalysts who want the deepest visual exploration available.Creator $75/user/month, billed annually
Looker StudioReporting on Google Ads, Analytics and Sheets with no budget.Free
FlitchTeams who want dashboards, reports, apps and workflows built from a description.Free tier, then $20/month
SigmaFinance and operations people who work in spreadsheets.No public pricing
ThoughtSpotTeams who would rather people asked a question than opened a dashboard.Essentials from $25/user/month, billed annually
MetabaseTeams with an engineer who can run a container.Free self-hosted, or $90/month + $6/user
Apache SupersetEngineering teams who want to run their own BI layer.Free, self-hosted
HexData teams who work in notebooks.Free Community tier, then $36/editor/month

Enterprise BI

Power BI

Power BI

Free account, then $14/user/month paid yearly
Best for.
Companies already running Microsoft 365, Azure or Fabric.
Where it wins.
$14 per user per month paid yearly, the cheapest enterprise BI seat there is.
What it costs you.
Copilot is not in that price. It needs a Fabric capacity from $262.80 a month on top of seats, and Desktop is Windows only.
Read the full comparison

Tableau

Creator $75/user/month, billed annually
Best for.
Analysts who want the deepest visual exploration available.
Where it wins.
More control over how a chart looks and behaves than anything else here.
What it costs you.
$75 per creator per month billed annually, and it expects someone whose job is building views.
Read the full comparison

Looker Studio

Free
Best for.
Reporting on Google Ads, Analytics and Sheets with no budget.
Where it wins.
Free, with Google sources connected in a couple of clicks.
What it costs you.
Thin modelling, and it slows badly once the data is large or needs joining.
ThoughtSpot

ThoughtSpot

Essentials from $25/user/month, billed annually
Best for.
Teams who would rather people asked a question than opened a dashboard.
Where it wins.
Search straight onto a live warehouse, with an agent over it.
What it costs you.
From $25 per user per month billed annually, and it needs a modelled warehouse first.
Read the full comparison

Built around AI

Flitch

Flitch

Free tier, then $20/month
Best for.
Teams who want dashboards, reports, apps and workflows built from a description.
Where it wins.
AI on every plan including the free one, and every chart shows the dataset, columns and calculation behind it.
What it costs you.
SCIM and running it in your own cloud need an Enterprise agreement, and it does not build an app with its own auth and payments.

Spreadsheet-style

Sigma

Sigma

No public pricing
Best for.
Finance and operations people who work in spreadsheets.
Where it wins.
A spreadsheet over the warehouse, and it writes back.
What it costs you.
No public pricing, and you need a cloud warehouse before you start.
Read the full comparison

Open source and developer-first

Metabase

Free self-hosted, or $90/month + $6/user
Best for.
Teams with an engineer who can run a container.
Where it wins.
Free under AGPL v3 for unlimited users if you self-host. That is a genuine advantage.
What it costs you.
Someone deploys it, upgrades it and keeps it up, and AI sits on the paid tiers.
Read the full comparison

Apache Superset

Free, self-hosted
Best for.
Engineering teams who want to run their own BI layer.
Where it wins.
Free, and running at scale inside large companies.
What it costs you.
Standing it up and keeping it current is a real job, and it assumes SQL.
Hex

Hex

Free Community tier, then $36/editor/month
Best for.
Data teams who work in notebooks.
Where it wins.
Python, SQL and no-code cells in one document.
What it costs you.
$36 per editor per month after the free tier, and it is built for analysts rather than for the people reading the output.
Read the full comparison

When to stay with Looker

  • You have data engineers who will write LookML and keep it current.
  • One governed definition of every metric across a large company is the requirement.
  • You are committed to Google Cloud and want BI bought alongside it.
  • Access rules belong in the model rather than one filter per dataset.

Questions

What does Looker cost?

Google publishes no price. It is a platform fee plus per-user seats on a 1, 2 or 3 year term, quoted by sales. Standard covers up to 50 internal users and an instance comes with 10 Standard and 2 Developer seats.

Is Looker Studio the same thing as Looker?

No, and the shared name causes real confusion. Looker Studio is the free reporting tool that used to be Data Studio. Looker is the governed modelling platform you buy from Google Cloud. Neither is the cheap version of the other.

Can I keep a semantic layer without writing LookML?

Yes. Flitch has relationships and calculations you set up without code, Sigma has datasets and models, and Basedash has governed metrics. None of them put the model in version control the way LookML does, which is the trade.

Do my LookML models move with me?

No. LookML is Looker’s own language and nothing else reads it. What carries over is the SQL behind each view and the definition of each metric, so export those while you still have the instance.

Which alternative suits a company already on BigQuery?

Looker Studio is free and reads BigQuery directly, though it slows on large joins. Flitch, Power BI, Tableau, Metabase and Sigma all connect to BigQuery as well, so the warehouse is rarely what decides it.