Keyboard-first terminal UI
Navigate, search, sort, and filter without leaving the keyboard.
- Consistent shortcuts across dialogs
- High-FPS rendering path
- Minimal latency input handling
DataTUI
A Terminal UI for Data.
A fast, keyboard-first terminal UI for exploring CSV/TSV, Excel, SQLite, Parquet, and JSON with tabs, sorting, filtering, SQL (Polars), and more.
Features
Broad import support
Guided dialogs for CSV/TSV, Excel, SQLite, Parquet, and JSON with type-aware options.
- Delimiter, header, quoting for CSV/TSV
- Sheet/table selection for Excel/SQLite
- Native Parquet/JSON flows
Multi-column sort
Sort by multiple columns with clear header indicators and toggle behavior.
- Ascending/descending toggles
- Stable multi-key sorting
- Visual arrows and order badges
Custom filter builder
Compose complex filters without writing code. Apply and clear against your base dataset.
- Variety of operators
- Export/Import filters
- Nested AND/OR conditions
SQL on your data
Run SQL via Polars SQLContext, including custom UDFs for specialized logic.
- Register and use UDFs
- Blend with non-SQL workflows
- Fast in-memory execution
JMESPath transforms
Apply JMES expressions and add derived columns with a runtime preloaded with custom functions.
- Built-in plus custom functions
- Safe expression evaluation
- Composable with filters
Find and Find All
Regex/whole-word/case options, forward/back, wrap, and a contextual results list to jump to matches.
- Regex and normal modes
- Match context preview
- Jump to selection
DataFrame details
Inspect schema, shapes, and types at a glance with a dedicated details view.
- Rows, columns, and dtypes summary
- Per-column types and metadata
- Quick copy of schema
- Heatmap
Column width control
Auto-expand columns by content or set manual widths. Hide/show columns as needed.
- Auto or manual sizing
- Per-column overrides
- Hide/show visibility
Column operations
Reorder columns, cast types, and perform advanced operations including clustering flows.
- Generate embeddings
- Clustering options (k-means, DBSCAN)
- Principal Component Analysis
More features
Tabbed data views
Open multiple datasets at once and manage them with ease.
- Independent state per tab
- Quick switch and reordering
- Join multiple datasets with SQL
Workspace persistence
Resume where you left off with saved state and snapshots.
- State and settings files
- Per-tab Parquet snapshots
- Instant startup on large datasets
Polars-powered engine
Efficient transformations and fast rendering on large data.
- Lazy query plans
- Efficient collection to views
- Scales to millions of rows
AI embeddings in SQL (optional)
Generate vector embeddings for text columns via an external provider.
- Opt-in integration
- Batch unique value optimization
- Great for semantic workflows
Export data
Export current view or dataset with guided dialogs and defaults.
- Choose format and scope
- Preview before writing
- Fast I/O paths
Project settings & data hub
Configure your workspace and manage sources from one place.
- Project defaults and paths
- Source browsing and selection
- One-click tab creation