The best way to control Claude Code, Codex and more from any computer or phone
docs(import): correct conversation size and batch limits (#23226)
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Victor Vogel committed
Oct 21, 2026 at 20:35 UTC
49e97d48f9742ba4c5ab79efc618fbf9066523ec
1 file changed
+4
-2
docs/user/welcome-wizard.md
+4
-2
@@ -83,10 +83,12 @@ It omits tool activity and attachments. For Codex, it omits generated setup
83
context only when a canonical user event and a valid shared turn ID identify the
84
same user turn. Ambiguous legacy or response-only context stays in the imported
85
conversation so T3 Code does not remove user text. It reads one conversation at
86
-a time and skips files larger than 16 MiB. It ignores malformed records and skips
86
+a time and skips files larger than 4 GiB. It also skips a conversation when the
87
+message text and metadata it keeps need more than 32 MiB of memory. Tool output
88
+does not count toward that limit. It ignores malformed records and skips
89
unreadable or unparseable conversations.
90
89
-Each import attempt reads up to 100 conversation files and 64 MiB per project,
91
+Each import attempt reads up to 100 conversation files and 4 GiB per project,
92
with up to 100,000 input records. Run import again to continue a large batch.
93
Completed conversations are not imported again. You can continue without the
94
remaining history.