Repository navigation
@ai-sdk/gateway@3.0.213
·
2881 commits
to main
since this release
Patch Changes
- 0a2b345: feat(google): add
gemini-4-argonmodel id and long decode support forgenerateContent: 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.generateTextwith the Interactions API now pollsbackground: truemodel calls until they finish. - 92ef6a5: Backport: chore(provider/gateway): update gateway model settings files
- Updated dependencies [604b4d3]
- @ai-sdk/provider-utils@4.0.58