High Performance Local Data Lake · Open Source

The Sublime Text
of Data

Open 5GB files in under 500ms. Query 46.7 million rows with SQL. No Python. No Spark. No cloud. Just your data, instantly.

468ms
Time to Data (5GB)
46.7M
Rows Rendered
258MB
Peak RAM (5GB)
641ms
Full-Table SQL
parq-bench — enhanced_dataset.parquet
# Open a 5GB IoT security dataset
▸ OPEN D:\datasets\enhanced_dataset.parquet
✓ Schema: 69 columns · Rows: 46,686,579 · Size: 5.00 GB
✓ Data visible: 468ms · Full grid ready: 637ms
 
# Full-table aggregation across 46.7M rows
▸ SELECT COUNT(*), AVG(col1), SUM(col2), ... FROM data -- 12-column aggregation
✓ 1 row in 641ms · Scanned 46.7M rows · Peak RSS: 140MB
✓ CTE + Window query: 3,214ms · Peak RSS: 181MB · No spill-to-disk
Your Local Data Lake

Built for files too big for everything else

⚡
DuckDB Engine
Embedded OLAP SQL engine memory-maps files directly from NVMe. Full SQL support with vectorized execution and predicate pushdown.
📈
Perspective.js Renderer
WebGL-powered datagrid renders millions of rows at 60 FPS. Pivot tables, charts, and cross-filtering built in.
📂
Multi-File Workspace
Register files as virtual tables. JOIN across datasets, compare schemas, and work with Hive-partitioned directories via glob patterns.
🔒
100% Local
Your data never leaves your machine. No accounts, no telemetry, no cloud. Runs on Tauri with a minimal footprint.
🛠
Format Transcoding
Convert between Parquet, CSV, JSON, and Arrow IPC. Fast export of filtered selections without loading full datasets into memory.
🔍
Schema Diff
Compare column layouts across files instantly. Spot added, removed, or changed columns before your pipeline breaks in production.

The gap between crashing and overkill

Excel Pandas Spark Parq-Bench
Open 10GB file ✕ Crashes ~45s + 20GB RAM ~60s startup ✓ <1s
SQL queries ✕ Via pandasql ✓ ✓ DuckDB
Parquet support ✕ ✓ ✓ ✓ Native
No install / setup ✓ ✕ Python env ✕ JVM + cluster ✓ Single binary
RAM for 5GB file — ~20GB ~8GB <260MB
Runs offline ✓ ✓ Mostly ✓ Always

Stop waiting. Start seeing.

Open your first 5GB file in under 500ms. Free and open source.

↓ Download for Windows ★ Star on GitHub