Skip to content

Conversation

snipeship
Copy link
Contributor

@snipeship snipeship commented Aug 19, 2025

this is just a proposal to make the process of enabling reasoning on Anthropic models not a pita.

the current implementation can cause max tokens errors when budgetTokens is set, forcing users to debug token limits, etc.

this change fixes that by adjusting the max output tokens automatically.

closes #2055

@snipeship snipeship force-pushed the hotfix/budget-output-tokens branch from bc85f92 to 0b5200b Compare September 2, 2025 12:04
@rekram1-node rekram1-node force-pushed the hotfix/budget-output-tokens branch from 0b5200b to 2ce4744 Compare September 7, 2025 18:03
@rekram1-node rekram1-node merged commit 4654fb8 into sst:dev Sep 7, 2025
1 check passed
lacymorrow pushed a commit to lacymorrow/lash that referenced this pull request Sep 8, 2025
aemr3 pushed a commit to aemr3/opencode that referenced this pull request Sep 18, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Bug: incorrect max_tokens calculation when "thinking" is enabled & "budgetTokens" is set for Anthropic models
2 participants