D
DuckDB MCP
45
Stack Score
Editorial pick✓ Verified workingMCP server
DuckDB MCP
Run analytical SQL on local + remote files via DuckDB.
Best for
Best for analysts and data engineers doing local-first or cross-source SQL with Claude.
At a glance
- Pricing
- open-source
- Setup time
- 15 minutes
- Learning curve
- easy
- Last verified
- 5/25/2026
What we love
- ✓Zero-setup analytical SQL
- ✓Faster than Pandas on disk-resident data
- ✓Cross-source queries
Where it falls short
- −Single-process — not multi-user
- −Persistence files need care
- −Best for analyst use cases
Pricing
✓ Pricing verified Jun 14, 20261 tier
Open-source
$0
perpetual
- ✓MIT
Overview
MCP server for DuckDB — query Parquet/CSV/JSON files directly, run analytical SQL, attach Postgres/MySQL/SQLite. The Swiss-army analytical engine, exposed to Claude.
Key features
- ●Query files directly
- ●Attach external DBs
- ●Embedded execution
- ●Parquet + CSV + JSON
- ●Lightning fast on local
Best use cases
Local data exploration
Have Claude write DuckDB SQL across local CSVs and Parquet for fast analysis.
Cross-source join
Query and join Postgres + S3 Parquet + local CSV from one DuckDB session.
Integrations
Similar tools
See all analytics →One AI Stack pick every week.
The tool we’d pay for this week, why we’d pay for it, and what we’d skip. No promo spam.