---
author: lukasmasuch
created: 2025-12-03
---

# Disable Data Export for `st.dataframe`

## Summary

Add a `client.disableDataExport` config option that controls whether data export functionality
(CSV download and clipboard copy) is disabled in `st.dataframe` and other components that
provide data export capabilities (e.g., Vega charts via the table view).

## Problem

### User Requests

- [#8402](https://github.com/streamlit/streamlit/issues/8402) — Possibility to disable the
  "Download as CSV" button
- [#11358](https://github.com/streamlit/streamlit/issues/11358) — Disable copying data from a
  dataframe

### Pain Points

Some users need to display sensitive data in dataframes while preventing easy bulk export.
Currently, `st.dataframe` provides:

- A toolbar button to download data as CSV
- Keyboard shortcut (Ctrl/Cmd+C) to copy selected cells to clipboard

For internal dashboards or applications displaying confidential information, users want to
restrict these export capabilities to make it less convenient to extract large amounts of data.

**Note:** This is a convenience feature, not a security control. Technically skilled users
can still extract data from the frontend.

## Proposal

### Config Option

```toml
# .streamlit/config.toml
[client]
disableDataExport = true  # default: false
```

Or via command line:

```bash
streamlit run app.py --client.disableDataExport=true
```

Or programmatically:

```python
import streamlit as st

st.set_option("client.disableDataExport", True)
```

**Config Option:**

| Option | Type | Default | Description |
|--------|------|---------|-------------|
| `client.disableDataExport` | `bool` | `false` | Whether to disable data export (CSV download and copying to clipboard) in components that support it |

### Behavior

When `client.disableDataExport = true`:

- **Download button**: Hidden from the toolbar in `st.dataframe`, `st.data_editor`, and chart
  components
- **Copy to clipboard** (`st.dataframe`): Completely disabled — pressing Ctrl/Cmd+C while a
  dataframe is focused will not copy any cell data to the clipboard (applies to both single
  and multi-cell selections)
- **Copy to clipboard** (`st.data_editor`): Remains enabled because `st.data_editor` relies on
  copy/paste for core editing workflows (e.g., pasting data into cells, copying rows)
- **Chart table view**: CSV export disabled in Vega chart table views (e.g., `st.altair_chart`,
  `st.vega_lite_chart`)
- **Other features**: Unaffected (search, fullscreen, sorting, selections all work normally)

This applies to all components with data export capabilities in the app.

### Examples

#### Disable Export App-Wide

```toml
# .streamlit/config.toml
[client]
disableDataExport = true
```

```python
import pandas as pd
import streamlit as st

df = pd.DataFrame({"Name": ["Alice", "Bob"], "Salary": [100000, 120000]})

# Export is disabled for all dataframes
st.dataframe(df)
```

#### Enterprise Deployment

Organizations can enforce this setting across all deployed apps via command-line flags
or environment variables, preventing accidental data leakage:

```bash
streamlit run app.py --client.disableDataExport=true
```

## Out of Scope (Future Work)

### Per-Dataframe Parameter

A per-dataframe `disable_export` parameter could be added later if users need granular control
(similar to how `st.file_uploader` has both `server.maxUploadSize` config and a per-widget
`max_upload_size` parameter):

```python
# Future: override config for specific dataframes
st.dataframe(df, disable_export=False)  # re-enable for this dataframe
```

This would allow the config option to set the default while still permitting exceptions.
Adding this later based on user feedback avoids premature API expansion.

## Alternatives Considered

### Per-Dataframe Parameter Only

Adding `disable_export: bool = False` directly to `st.dataframe`.

**Rejected because:**

- `st.dataframe` already has many parameters; adding more for niche use cases adds clutter
- Most use cases want to disable export for all dataframes in an app, not selectively
- Easy to forget setting the parameter on one dataframe, creating inconsistent behavior
- Cannot be enforced at the deployment/organization level

## Checklist

| Item                         | ✅ or comment |
|------------------------------|---------------|
| Works on SiS, Cloud, etc?    | ✅ |
| No breaking API changes      | ✅ |
| No new dependencies          | ✅ |
| Metrics collected            | ✅ |
| Any security/legal impact?   | ✅ Not a security feature; documented as convenience only |
| Any docs changes needed?     | ✅ Clarify this doesn't prevent determined data extraction |