[0]Welcome

BigQuery,
with everything you wished it had.

Querylab.io is a SQL IDE for BigQuery, in your browser or as a desktop app.

Free. The Playground needs no sign-in.

[1]Events export
-- cost shown before you run
SELECT *
FROM `ace-analytics.warehouse.events`
WHERE event_date BETWEEN '2019-01-01' AND '2024-12-31'
[3]What it does
Write
Completions, go to definition and hover docs that know your schema
Check
Errors and warnings before you run, each with a fix
Run
On-demand or slots, whichever costs less
Debug
Click a step of the query plan to see the SQL behind it
Limits
GB caps on every query, including the ones Claude Code runs
[4]Get started

Where to start

Playground
The full IDE on a sample project. Open it without signing in.
Your projects
Sign in with Google and query the projects you already have. BigQuery returns rows straight to your browser. The desktop app keeps your sign-in in the OS keychain.
Learn
Learn BigQuery is a free course of 76 lessons. They run in your browser and need no account.
Querylab.io
BigQuery, with everything you wished it had.

Querylab.io is a SQL IDE for BigQuery, in your browser or as a desktop app.

-- cost shown before you run
SELECT *
FROM `ace-analytics.warehouse.events`
-- 80 TiB | ~$500

Free. The Playground needs no sign-in.

Cost control

BigQuery charges by the byte. We show it in dollars as you type.

PREVIEW

See the cost while you write

Every query, estimated as you type: dollars, not bytes. The per-CTE breakdown points at the expensive part.

BUDGET

Budgets per user, not per project

BigQuery caps every user with the same quota. Querylab.io sets a limit per person: 10 TiB a day for your senior analyst, 500 GiB for the new hire. Over-budget queries stop before they bill.

SAMPLE

Test on a sample

Set a Tablesample percentage and the query reads a sample of each table. BigQuery can't sample views, and a table up to about 168 MiB is one block, so those are read in full. Querylab.io prices the sample and the full query, so you see what you saved.

vs. the console

The Console manages your cloud. We write SQL.

Query cost
Bytes estimate, top-right of the editor.
✓Dollars as you type, per CTE.
Spend limits
Project-level quotas, same for every user.
✓Budgets per user, query or day.
Cost mistakes
No warning for a missing partition filter.
✓Underlined as you type, like a typo.
Cheap iteration
Add TABLESAMPLE to each table by hand.
✓Set one percentage. Querylab.io samples the tables it can and prices the savings.
Debugging a CTE
Comment out the rest of the query and re-run.
✓Click the CTE to run or price just that part.
Slots vs on-demand
After the run: job stats → slot-ms × your rate.
✓On-demand priced, slots predicted. Auto mode runs each query the cheaper way.
Autocomplete
Keywords and table names.
✓Schema- & scope-aware suggestions.
Focus
One tab of the full GCP Console.
✓Query, results, schema. ⌘K for the rest.
What you get

The features you'll use every day.

Per-CTE cost breakdown

See which part of a query burns bytes, before you run it.

Real-schema completions

Tables, columns and values from your actual data, scope-aware.

Budget limits

Cap spend per query, user or day. Expensive queries stop first.

On-demand vs reservations

Both pricing models priced on the same query, before you run it.

Prism AI

Write, fix and explain SQL with an agent that knows your warehouse.

All features

Everything Querylab.io does.

Writing SQL

Completions that know the clause · Real values in WHERE · Ghost text without AI · Checks as you type · Pipe syntax · Graph queries · dbt and Dataform · Go to definition, rename · Formatting

Cost

Budgets and limits · Auto routing · Tablesample · On-demand or slots · AI function costs · LIMIT hints

Understanding a query

Run one CTE · Column lineage · Insights panel · Execution plan · Charts in the SQL

AI

Prism AI · MCP server · AI ghost text · Inline edit

Data

A table in one click · Query results for free · Data compare · Import · Export · Time travel · Storage costs · GA4 exports · Bulk operations · DDL from the tree

Workspace

Notebooks · Query history · Share links · Results grid · Tabs and split view · Command palette · Notifications

Desktop app

Local BigQuery API · Your own AI subscription · gcloud sign-in

Learning

Playground · Learn BigQuery · BigQuery concepts

Cost-aware MCP

Your agent queries BigQuery through the limits you set.

The desktop app runs an MCP server on your machine. Claude Code, Codex, Cursor, Gemini CLI, Claude Desktop and VS Code connect with one button each, and every call they make goes through your caps, dataset rules and approvals.

A cap and a budget

Maximum GiB per query, plus a daily limit for AI tools alone. A query over either one is refused before BigQuery bills for it. On a reservation, a slot limit per query as well: the query is cancelled once it passes it, and the slot time it used is billed.

On-demand or slots, per query

The same query can cost several times more on one billing model than on the other. Auto routing prices both after the dry run and sends the query to the cheaper one.

Reads run, writes wait

A SELECT inside your limits runs without asking. INSERT, UPDATE, DELETE, MERGE and schema changes are off until you allow them, and then they still stop on a card that shows the statement and the table it changes.

The agent knows its own limits

get_limits reports the caps, today's usage and your dataset rules, so an agent plans a query it is allowed to run instead of finding out by being refused.

On macOS and Windows. The server runs on your machine.

Desktop app

Native on your desktop.

The same IDE without the browser. macOS, and Windows in beta.

Bring your own AI

Prism AI drives the agents already on your machine: Claude Code, Codex, Gemini CLI, Copilot, Cursor. Your subscriptions, your keys.

MCP server

Claude Code, Codex, Cursor and other AI tools query BigQuery through the limits you set.

Existing gcloud auth

Already logged in with gcloud? The app picks it up automatically. No extra sign-in.

Credentials stay local

BigQuery credentials live in your OS keychain, never on our servers.

Native hotkeys

No browser shortcuts getting in the way. No tab to lose.

Local BigQuery API

bq and the Python client run SQL on your computer instead of BigQuery. Nothing is billed.

Learn BigQuery

Learn BigQuery SQL, free in your browser.

76 lessons and 251 challenges on UK car safety tests and a Google Analytics 4 event export. Every answer is checked. No account needed.

Part 1: SQL foundations

Basics · Aggregation · Joins · Dates and strings · Subqueries and CTEs · Window functions · Arrays and structs

Part 2: Advanced BigQuery features

BigQuery for SQL users · Cost and optimization · Pipe syntax · Scripting · Writing data

Part 3: Analytics on real data

Events, sessions, funnels and retention on GA4 export data.

Part 4: Capstones

An open problem with no step-by-step guide.

Pricing

The editor is free.

No credit card and no trial. Sign in with Google and start.

What's includedprice
  • Full SQL editor$0.00
  • Unlimited queries$0.00
  • Cost preview, per CTE$0.00
  • Schema-aware completions$0.00
  • Budget limits, per-query caps$0.00
  • On-demand vs reservation pricing$0.00
  • Prism AI$0.00
  • Pipe syntax$0.00
  • Notebooks$0.00
  • Query history$0.00
  • Lineage$0.00
  • Import, export, data compare$0.00

The Playground needs no sign-in.

Your BigQuery billGoogle bills you for BigQuery as before. The cost calculator estimates that bill and how much Querylab.io could take off it.
Get started

Best on the big screen.

Querylab.io is a full SQL IDE, and it shines on a laptop. Send yourself the link and open it when you're back at your desk.

Just the link, straight to your inbox. Updates stay off unless you tick the box.

or jump straight in
Get started →