Docs
PlaygroundOpen app
Docs/What's new/Release notes
What's new

Release notes

Features, SQL support and improvements.

1.5Optimize, fix buttons and slot limits

Main features

  • Optimize — Where your BigQuery spend went, what to fix first, and slots against on-demand.
  • Fix buttons on errors — One-click fixes on a failed query’s error message and on its marker in the editor.
  • Slot limit per query — Caps a query’s slot time, in tabs, notebooks and connected AI tools.
  • Ghost text that learns — Learns aliases, joins, IS NULL and GROUP BY habits from your past queries.
  • Windows app (desktop) — Querylab.io IDE in the Microsoft Store, in beta. The Mac app is out of beta.

BigQuery syntax

  • REGEXP_EXTRACT_GROUPS: completions and hover for the fields of its STRUCT, named after your capture groups.
  • AI.TRANSLATE, AI.CHUNK_DOCUMENT and the new AI.AGG options.
  • Chained calls with a dotted function name, such as ('hello').(AI.TRANSLATE)('de').

Other updates

Query history is a list with a detail pane, with filters and keyboard access. Open results in a local tab to query a result in the browser at no cost. The lineage graph links to the editor. The Insights panel is organised around answers. Storage statistics price each table on its cheaper billing model. Choose which reservations Auto routing may use. Format applies to the SQL you paste. Bulk operations work on a phone and from the keyboard. Spaces are always on. On desktop, connected AI tools can read their limits and ask you to lift one for a call, the tray icon shows calls that wait for you, and the native menus reach every app feature.

Performance improvements, stability updates and bug fixes.

1.4Auto routing, typed parameters and the MCP tab

Main features

  • Auto routing — Auto picks on-demand or your reservation for each query, based on its past runs and its structure.
  • Typed query parameters — A -- querylab:param line gives @name a type and value, sent with every run and estimate.
  • MCP tab (desktop) — Activity, cost per AI tool, a full call log and pending approvals in one tab.
  • Completion keyboard on phones — Tap completions from a keyboard of SQL suggestions, or a bar above your keyboard.
  • Notifications — A system notification when a long query finishes, or when Prism AI is done or needs you.

BigQuery syntax

Other updates

Auto routing works in notebooks, per notebook and per cell. Edit parameter values in the status bar ⚙ panel, and Declare parameter adds a missing one. A LIMIT you wrote shows when it can cut the scan. Dataset DDL lists the authorizations a GRANT can’t express. On desktop, connected AI tools get access limits, dataset rules and a Write approval setting, and the MCP server starts the app when an AI tool needs it. Each Google account keeps its own local tabs and history, and Sign out deletes only that account’s data. Format says when it leaves a query as written.

Performance improvements, stability updates and bug fixes.

1.3.9Stability update

desktop only

  • Performance improvements, stability updates and bug fixes.

1.3.8BigQuery tools for AI clients

desktop only

  • Local MCP server: connected AI tools explore schemas, sample tables, validate SQL, estimate costs and run SELECT queries.
  • Performance improvements, stability updates and bug fixes.

1.3.5Local MCP server and BigQuery REST API

Main features

  • Local MCP server — Connect Claude Code, Claude Desktop, Codex, Gemini CLI, Cursor or VS Code to Querylab.io to work with BigQuery.
  • Local BigQuery REST API — Run SQL locally from bq, Python and other BigQuery clients through the desktop app’s BigQuery-compatible REST API.

BigQuery syntax

  • Identity columns: GENERATED ALWAYS AS IDENTITY and GENERATED BY DEFAULT AS IDENTITY.
  • Aggregate filters: WHERE inside aggregate calls, including COUNT(* WHERE ...).
  • OBJ.LIST: editor support for the table function and its output columns.
  • AI.CAUSAL_EFFECT: editor support for the table function and its output columns.
  • ML.METRICS: editor support for the table function and its output columns.
  • ML.CORRELATION: editor support for the table function and its output columns.

Other updates

Pipe conversion adds support for PIVOT, UNPIVOT and value tables. DATETIME and TIME results use BigQuery console formatting. Table DDL includes identity columns and constraints. Editor checks cover table definitions and table-function arguments.

Performance improvements, stability updates and bug fixes.

1.3Tab to expand, LIMIT suggestions

Main features

Performance improvements, stability updates and bug fixes.

1.2.9Pipe syntax support completed

Main features

  • Pipe syntax — Completions, hover, rename, go to definition, checks and lineage for pipe queries.
  • Column lineage — Trace struct fields, correlated subqueries, ARRAY subqueries and SELECT * EXCEPT.

BigQuery syntax

Other updates

Column types appear beside SELECT items. SQL notebook cells can be edited on iPhone and iPad.

1.2.7Spaces and project setup

Main features

  • Spaces — Group tabs and projects, with separate billing and visibility per Space.
  • Project setup — Connect a project on web or desktop, or start in the playground.

Other updates

Query results show estimated and actual bytes scanned.

Performance improvements, stability updates and bug fixes.

1.2.5Graph queries and charts

Main features

  • Graph queries — GQL completions, navigation and checks, with a graph view of results and SQL-to-GQL conversion.
  • Charts from SQL — Render charts with Querylab.io’s optional VISUALIZE clause.
  • dbt macros — Completions, hover and checks work with expanded macros.

BigQuery syntax

Other updates

Choose a default reservation for new query tabs and override it per query.

1.2.3Pipe syntax conversion and the Mac app

Main features

Performance improvements, stability updates and bug fixes.

1.2Playground and inline AI edits

Main features

  • Playground — Use the full editor on public BigQuery data without signing in.
  • Inline AI edits — Review suggested changes to the current query as a diff.
  • dbt and Dataform — Compile templates with inline cost estimates and editor checks.
  • Column lineage — Trace columns through CTEs, views and table functions.

BigQuery syntax

Other updates

Share notebooks by link, bind parameters to cells and copy cells between notebooks.

1.1.5Notebooks, compare, scheduled queries

Main features

  • Compare tab — Two queries or two schemas side by side with differences marked.
  • Scheduled queries — Create, edit and browse scheduled queries, with timezones and run history.
  • Notebooks — SQL, markdown and results in one document, with values passed between cells.
  • Prism AI tools — Schema exploration, query runs and comparisons, saved queries, choice of provider.
  • Import and export — Drag files in. Export to Excel, DuckDB, dbt models, Cloud Storage or the clipboard.
  • Insights panel — Query complexity, slot and reservation recommendations, prices in your currency.

1.1Prism AI

Main features

  • Prism AI — An assistant that works from your schema, in read-only, confirm or autonomous mode.
  • Ghost completions — The rest of the line suggested ahead of the cursor, including joins.
  • Spaces — Separate tabs, pinned tables and a colour theme per project.
  • Query sharing — A link that carries the SQL and opens it in a tab.
  • Value completions — Actual column values suggested in WHERE comparisons.
  • Google Sheets export — Results sent to a new or existing spreadsheet.
  • GA4 datasets — Recognised automatically, with field descriptions and 21 ready-made queries.

1.0First public release

Main features

  • Editor checks — Over twenty checks run as you type.
  • Cost breakdown — A per-column cost estimate for each query, with a 100 GB default ceiling.
  • Query history — Runs kept with cost, job details and the tables involved, reopened with their results.
  • Command palette and split view — Cmd+K for actions, tables and past queries. Tabs side by side.
  • Storage analysis — Storage cost per table, physical and logical billing, partitioning and clustering.
  • Settings — Formatting, completions, editor behaviour and privacy.

0.9Statement execution and grid rework

Main features

BigQuery syntax

Other updates

Function completions are grouped by namespace. Sharded tables appear as one entry; views have details and query tabs.

Performance improvements, stability updates and bug fixes.

0.7SQL parsing in the browser

Main features

  • SQL editor — Completions, hover and checks run as you type, including inside joins and subqueries.
  • Code navigation — Go to definitions and find references from the editor.
  • Pipe conversion — Convert standard SQL to pipe syntax.

BigQuery syntax

0.5Sign-in, formatting, dark theme

Main features

  • Google sign-in — A guided first run from sign-in to a selected project.
  • Demo mode — Public BigQuery data, no project required.
  • SQL formatting — Format a query with one action.
  • SQL generation — SELECT, CREATE and other statements from the schema tree.
  • Dark theme — For the editor, grid and dialogs, stored with your account.
  • Query tabs — Each tab keeps its own SQL and results.

0.3Column completions and cost estimates

Main features

  • Column completions — Columns suggested from the schemas of the tables you reference.
  • Results grid — Merged cells for repeated values, formatted dates and timestamps.
  • Cost estimates — An estimated price for each query before it runs.
  • AI query fixes — A corrected version of a query on request.
  • Browser install — Querylab.io can be installed and opened like a desktop app.

Performance improvements, stability updates and bug fixes.

0.1First working version

Main features

  • SQL editor — BigQuery syntax highlighting and table name completion.
  • Schema sidebar — Projects, datasets and tables in a sidebar.
  • Query results — Results below the editor, loaded page by page.
  • Background jobs — Queries run as BigQuery jobs and continue in the background.