Zed 1.17 Turns Your Code Editor Into an Interactive Data Viewer

The Zed editor just shipped version 1.17, and the headline feature turns the editor into a lightweight data viewer. Open any CSV, TSV, PSV, or SSV file and Zed now renders it as an interactive table you can actually work with, alongside Git and memory improvements aimed at people who spend the day inside large repos.

This is a weekly release from the Zed team, so nothing here is priced separately. Grab the update from the built-in updater or the download page, and the tabular preview kicks in automatically the next time you open a delimited file.

Spreadsheets, without leaving the editor

The centerpiece is a proper tabular preview for delimited data, available to all users. It handles CSV, TSV, PSV, and SSV files, with sortable and independently resizable columns, row filtering by selected values across one or more columns, a row-number gutter that stays pinned during horizontal scrolling, and right-click copy for cells and column headers.

That covers most of the reasons developers usually bounce out to a spreadsheet app or a Jupyter notebook when a teammate drops a data dump in the repo. A few concrete things you can now do without leaving Zed:

• Click a column header to sort, then drag its edge to resize.
• Select a value in a column and filter every row that matches, stacking filters across multiple columns.
• Keep the row-number gutter pinned while you scroll a wide table horizontally.
• Right-click any cell or header to copy it straight to the clipboard.

Big files, smaller footprint

Memory handling got a rework too. Opening a large file now reduces peak memory during loading by roughly the size of the file itself. A multi-hundred-megabyte log or dataset no longer briefly doubles Zed's resident memory while the buffer is being constructed, which is the kind of spike that used to trigger swap or OOM kills on smaller laptops. Editor rendering speed also improved in the same release.

Git panel and blame gutter get sharper

The Git Panel picks up two new stash actions, Stash Tracked and Stash Staged, giving you finer control than a blanket git stash. If you routinely stage a subset of hunks and want to shelve just that subset while you jump onto a hotfix, this is the button you were missing.

The blame gutter also gets two new commands, Blame Revision and Blame Previous Revision, which open the file as it existed at the commit under your cursor, or at its parent. That is the classic "who broke this and what did the line look like before" workflow, now one click away from the inline blame instead of a trip to the terminal.

The rest of the changelog

A few smaller items are worth flagging for anyone tuning their setup:

• Support for low reasoning effort on DeepSeek V4 Flash and V4 Pro, useful if you drive those models through Zed's agent and want cheaper, faster responses on routine edits.
• The JetBrains keymap gains CamelHump-style subword navigation for alt-left, alt-right, shift-alt-left, and shift-alt-right in editors.
• Markdown table column sizing is now based on content, so wide preview tables stop squashing narrow columns.
• Filtering for language server completions has improved, which should mean fewer irrelevant suggestions surfacing at the top of the list.
• The file_scan_exclusions setting now supports a "..." entry, letting custom exclusions extend the defaults instead of replacing them.

Who this is for

If your work touches CSV exports from a database, telemetry logs, or ML datasets, the table preview alone is a meaningful quality-of-life upgrade. You get sortable, filterable inspection of tabular data in the same window as your code, without invoking pandas or opening a separate viewer just to check whether a column has the expected cardinality.

For everyone else, 1.17 is a maintenance release with some Git polish and a real memory win on large files. The full changelog is on GitHub, and the update should already be waiting in your editor.