### What feature would you like to see? ChatGPT Web should support selecting multiple conversations and archiving them in one operation. ### Current behavior ChatGPT currently supports: - Archiving one conversation at a time. - Archiving **all** conversations from Data Controls. There is no supported way to archive an arbitrary subset of conversations in bulk. For users with hundreds or thousands of conversations, individually archiving old chats is impractical, while **Archive all chats**

[ChatGPT Web] Add multi-select / selective bulk archive for chats

Open 💬 1 comment Opened Aug 27, 2026 by yhtang

What feature would you like to see?

ChatGPT Web should support selecting multiple conversations and archiving them in one operation.

Current behavior

ChatGPT currently supports:

Archiving one conversation at a time.
Archiving all conversations from Data Controls.

There is no supported way to archive an arbitrary subset of conversations in bulk.

For users with hundreds or thousands of conversations, individually archiving old chats is impractical, while Archive all chats is too destructive as an organizational tool.

Requested behavior

Add multi-select to the chat/history interface, with at least:

Select multiple chats.
Archive selected.
Select all currently visible/filtered chats.
Clear confirmation showing how many chats will be archived.

Useful follow-ups could include filtering by age, project, pinned/starred state, or last activity before selecting chats.

Why this matters

The underlying archive capability already exists; the missing piece is bulk selection.

Without a supported selective bulk operation, power users are pushed toward browser automation or undocumented internal ChatGPT APIs to manage large histories. A first-party bulk archive UI would be safer, more reliable, and avoid users having to automate unsupported endpoints.

This request is specifically for ordinary ChatGPT conversations, including ChatGPT Web, rather than Codex-local session management.

#32070 — related batch/automatic archiving request for Codex sessions.
#37791 — broader request for bulk organization in the unified desktop app.

This request is narrower: first-class selective bulk archiving of ordinary ChatGPT conversations.

View original on GitHub ↗

1 Comment

whatthecodemean · 21 hours ago

I hit the same gap from a slightly different angle: selective bulk management becomes much more useful if it can also support a bounded cleanup workflow, not only manual multi-select.

For example, an assistant may identify 30 conversations that look obsolete or duplicated. If the user still has to open 30 menus and archive/delete each one, most of the automation value is lost at the final step.

A safe flow could be:

build a candidate set from search/filter or an assistant suggestion;
show the exact conversations and count before changing anything;
let the user approve that set once;
archive the set in one operation, with permanent delete as a separate, stronger confirmation.

Archive is the better default because it is reversible. Delete should stay explicit. Files kept in ChatGPT Library should remain a separate lifecycle, so cleaning conversation history does not implicitly remove retained files.

The important primitive is not unrestricted agent deletion; it is a user-approved batch operation over a visible, bounded set. That would solve the practical case where dozens of clearly stale conversations have already been identified without pushing users toward browser automation or undocumented endpoints.