LakeDB Beta 6.0

LakeDB

PostgreSQL joins LakeDB natively.

Connect, explore, edit, design, export, operate and use review-first AI through PostgreSQL-specific runtime behavior.

Download Beta 6.0 Explore Beta 6

Windows x64 macOS Apple Silicon Linux x64

Native PostgreSQL runtime Safe typed editing Database tools QuerIA + AI Fix

01 / Beta 6.0

PostgreSQL behavior, end to end.

LakeDB uses PostgreSQL-specific connections, catalogs, quoting, types and plans instead of translating through another engine.

LakeDB Beta 6.0 native PostgreSQL capabilities

Connect securely

Use direct or SSH routes, PostgreSQL SSL modes, client certificates, read-only sessions and explicit transactions.

Work with the catalog

Browse schemas, materialized views, overloaded functions, sequences, types, keys, indexes and constraints.

Review every change

Edit with stable identities, inspect plans, design and transfer tables, or ask QuerIA and AI Fix without automatic execution.

02 / QuerIA

Direct when it is simple. Investigative when it is not.

QuerIA is a line-based query document beside your SQL tabs. Use Normal generation for focused requests or Agentic generation when the answer may require more tables, relationships and index review.

Choose

Use Normal generation for a direct draft or Agentic generation for deeper investigation.

Inspect

The agent can read relevant structures, keys and indexes across accessible databases.

Validate

SELECT plans are reviewed with EXPLAIN and their index signals are shown beside the SQL.

Run locally

The complete statement stays visible. Only you can execute it through the normal connection.

02 / Real examples

Beyond a one-table SELECT.

Beta 4.5 combines reviewable SQL with local result filters, draggable result columns and safer everyday SQL editing.

Analysis

“Show the 10 users who placed the most bets today, including bet count and total stake.”

SELECT u.id, u.email,
       COUNT(b.id) AS bet_count,
       SUM(b.stake) AS total_stake
FROM users AS u
JOIN users_bets AS b ON b.user_id = u.id
WHERE b.created_at >= CURRENT_DATE()
  AND b.created_at < CURRENT_DATE() + INTERVAL 1 DAY
GROUP BY u.id, u.email
ORDER BY total_stake DESC
LIMIT 10;

Controlled write

“Prepare an insert for a new user with email ada@example.com.”

INSERT INTO users (email)
VALUES ('ada@example.com');

Schema design

“Create a user_segments table with a unique name, description and timestamps.”

CREATE TABLE user_segments (
  id BIGINT UNSIGNED NOT NULL AUTO_INCREMENT,
  name VARCHAR(120) NOT NULL,
  description TEXT NULL,
  created_at TIMESTAMP NOT NULL DEFAULT CURRENT_TIMESTAMP,
  updated_at TIMESTAMP NOT NULL DEFAULT CURRENT_TIMESTAMP
    ON UPDATE CURRENT_TIMESTAMP,
  PRIMARY KEY (id),
  UNIQUE KEY uq_user_segments_name (name)
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4;

Focused update

“Update the user with id 7 and change the email to ada.lovelace@example.com.”

UPDATE users
SET email = 'ada.lovelace@example.com'
WHERE id = 7;

Schema evolution

“Alter the users table to add an optional last_login_at timestamp.”

ALTER TABLE users
ADD COLUMN last_login_at TIMESTAMP NULL;

Multiple independent SQL workspaces in LakeDB — **The complete client remains underneath QuerIA.**Every connection owns its SQL tabs, table tabs, active schema and restored workspace state.

03 / Complete SQL client

AI joins the product. It does not replace it.

LakeDB still gives every connection its own editors, object explorer, result tabs and restored workspace.

Schema-aware editorTables, aliases, columns, primary keys and index predicates
Large data gridsSearch, filters, typed editing, conflict checks and rollback
Database operationsExplain, transactions, backup, restore, compare and migrate
Independent workspacesLocal, staging and production contexts never blur together

04 / Compare

One complete workflow.
Fewer compromises.

LakeDB supports MySQL, MariaDB, PostgreSQL and SQLite through explicit engine capabilities rather than pretending every database offers the same workflow. This comparison highlights where its review-first approach differs from four established clients.

Included Limited, paid or a different workflow No comparable built-in workflow found No built-in AI / not applicable

Capability	LakeDB	DBeaver	TablePlus	MySQL Workbench	HeidiSQL
Connections, workspaces and core SQL
PK and index-aware completion hints	✓	◐	×	×	×
Restored multi-connection workspaces	✓	✓	✓	◐	◐
Environment labels and production safeguards	✓	✓	✓	×	×
Local syntax diagnostics before execution	✓	✓	×	◐	×
Typed staged edits with conflict detection and rollback	✓	◐	◐	◐	◐
Editor, data, database operations and recovery
Large-value JSON and HTML validation	✓	◐	◐	×	◐
Guided backup and restore with automatic recovery dump	✓	◐	◐	◐	◐
Schema compare and generated migration plan	✓	◐	×	✓	×
Natural-language SQL and trust boundary
Explicit AI error correction with complete SQL review	✓	◐	◐	—	—
Dedicated line-based natural-language documents	✓	×	×	—	—
Minimum useful schema grounding	✓	◐	◐	—	—
Agentic cross-database relationship and index inspection	✓	◐	×	—	—
SELECT, DML and DDL with visible review	✓	◐	◐	—	—
AI never receives row data or query results	✓	◐	✓	—	—
Explicit opt-in with no retained questions, context, SQL or schema names	✓	×	◐	—	—

Fourteen reasons to consider LakeDB. Shared capabilities such as multiple connections, SSH/SSL, normal autocomplete, editable grids and visual table editing are intentionally omitted. A partial mark covers paid tiers, plug-ins or a materially different workflow. A dash means the comparison is not applicable because the client has no built-in AI.

Sources: DBeaver SQL editor, DBeaver error correction, TablePlus documentation, MySQL Workbench syntax checker and HeidiSQL features and documentation.

05 / Explicit opt-in

No activation.
No service connection.

QuerIA starts disabled. You see the privacy notice and beta terms before LakeDB can contact the optional service. You can disable it again at any time, which stops service requests and requires fresh consent before reactivation.

LakeDB Beta 4.0 Beta Tester request workflow — **Real workflows shape QuerIA.**Request Beta Tester access from Settings → Account and tell us what you want to test.

06 / Beta Tester

Test the relationships that matter in real work.

Add an alias, email and a short use case inside the application. No password or email code is required.

Apply in LakeDBSettings → Account → Request Beta Tester
Describe one workflowNo credentials, customer data or confidential SQL
Report what went wrongTables, relationships, indexes or generated SQL

07 / QuerIA roadmap

From direct SQL
to bounded investigation.

LakeDB roadmap from the SQL foundation through QuerIA and native PostgreSQL toward 1.0 — **Product stages, not a release calendar.**Beta 6 adds native PostgreSQL behavior across the complete LakeDB workflow on the road to a trusted 1.0 release.

Read the complete roadmap ↗

08 / Download

Try the final public beta line.

Download Beta 6.0

QuerIA includes a free beta allowance. The complete SQL client remains usable when QuerIA is disabled, unavailable or has no remaining allowance.

Windows x64Installer and portable build↗ macOS Apple SiliconDMG and ZIP↗ Linux x64AppImage and Debian package↗

Independent development

Help LakeDB reach 1.0.

Patreon support funds infrastructure, signing and the work needed to turn this beta into a trusted release.

Support LakeDB on Patreon