You'll noticeTier: how much it should matter to you
1Useful: my rating, 1 to 5
1Signal: worth watching, 1 to 5
API RequestsArea: what it touches
Bug FixesKind: in v2.1.299,
Bug FixesSection of the release
What
When a reply from Claude is cut off because it reached the maximum output length, Claude Code shows an error. Before, that error always suggested raising CLAUDE_CODE_MAX_OUTPUT_TOKENS. For requests sent without streaming (where the reply arrives all at once rather than piece by piece), the message now says that this environment variable cannot raise the limit.
Why
The old advice was wrong for non-streaming requests, so following it did nothing. You will no longer be pointed at a setting that cannot help.
The entry above is what we published on the day. These lines were added later, as Anthropic's own pages caught up, and they sit beside the original rather than replacing it.
Confirmed sinceAnthropic's documentation has since written up CLAUDE_CODE_MAX_OUTPUT_TOKENS, on Error reference.API Error: Claude's response exceeded the ... output token maximum. To configure this behavior, set the CLAUDE_CODE_MAX_OUTPUT_TOKENS environment variable.errorssee the edit
How sure we are
One source agreesOne thing we can check says the same as this entry.
Anthropic's documentation agreesAnthropic's documentation has since written up CLAUDE_CODE_MAX_OUTPUT_TOKENS, on Error reference.
