Repository navigation
@ai-sdk/gateway@2.0.166
·
4728 commits
to main
since this release
Patch Changes
- 7c28815: feat(google): add
gemini-4-argonmodel id and long decode support: requests that reach the per-request cap (CONTINUATION) are continued automatically with usage summed across requests and the first request's prompt token count exposed asproviderMetadata.google.initialPromptTokenCount, and thecontinuationTokenprovider option / provider metadata resume responses that used up theirmaxOutputTokensbudget - 66941b0: Backport: chore(provider/gateway): update gateway model settings files