chore(deps): bump @ai-sdk/anthropic from 2.0.89 to 4.0.21

[dependabot]

Bumps @ai-sdk/anthropic from 2.0.89 to 4.0.21.

Release notes

Sourced from @​ai-sdk/anthropic's releases.

@​ai-sdk/provider@​4.0.8

Patch Changes

• 591d25b: feat: add batch completion webhooks. experimental_startTextBatch accepts a webhookUrl, and the gateway provider registers it through the batch callbackUrl contract and exports typed async-job metadata. Direct Anthropic and OpenAI batch providers return an unsupported warning when the option is provided.

@​ai-sdk/react@​3.0.269

Patch Changes

• Updated dependencies [a1bafb1]
• Updated dependencies [5642849]
• Updated dependencies [3cb50fc]
  • ai@6.0.266
  • @​ai-sdk/provider-utils@​4.0.48

@​ai-sdk/vue@​3.0.266

Patch Changes

• Updated dependencies [a1bafb1]
• Updated dependencies [5642849]
• Updated dependencies [3cb50fc]
  • ai@6.0.266
  • @​ai-sdk/provider-utils@​4.0.48

@​ai-sdk/gateway@​3.0.181

Patch Changes

• Updated dependencies [5642849]
  • @​ai-sdk/provider-utils@​4.0.48

@​ai-sdk/xai@​3.0.126

Patch Changes

• Updated dependencies [5642849]
  • @​ai-sdk/provider-utils@​4.0.48
  • @​ai-sdk/openai-compatible@​2.0.72

@​ai-sdk/google@​3.0.113

Patch Changes

• Updated dependencies [5642849]
  • @​ai-sdk/provider-utils@​4.0.48

@​ai-sdk/openai@​3.0.101

Patch Changes

• 313a441: Prevent negative text output token counts when providers report reasoning tokens. Perplexity reasoning tokens are now treated as separate from completion tokens.
• Updated dependencies [5642849]
  • @​ai-sdk/provider-utils@​4.0.48

@​ai-sdk/langchain@​3.0.79

... (truncated)

Changelog

Sourced from @​ai-sdk/anthropic's changelog.

4.0.21

Patch Changes

• e29788d: fix(anthropic): report thinking tokens as reasoning token usage

4.0.20

Patch Changes

• cbdc990: feat (provider/anthropic): support fallbacks 'default' mode, which routes safety classifier refusals to Anthropic's recommended fallback model (adds the server-side-fallback-2026-07-01 beta automatically)
• cbdc990: feat (provider/anthropic): support mid-conversation tool changes via the toolChanges system message provider option, emitting tool_addition/tool_removal content blocks and the mid-conversation-tool-changes-2026-07-01 beta
• cbdc990: feat (provider/anthropic): add claude-opus-5 model id with frontier-tier capabilities (128k output tokens, structured output, adaptive thinking, xhigh effort, sampling parameter rejection, thinking-disabled only at effort high or below)

4.0.19

Patch Changes

• 01a596a: fix (provider/anthropic): use current-generation capability defaults for unrecognized Claude model IDs while retaining conservative defaults for legacy Claude and non-Claude models.

4.0.18

Patch Changes

• 97de198: Warn when an unknown model uses the default 4096 max output token limit.

4.0.17

Patch Changes

• b72fc7c: fix(amazon-bedrock): sanitize unsupported JSON Schema constraints in native Anthropic structured output
• 9218ebe: fix(provider/anthropic): warn when parallel tool use is requested with JSON tool structured output
• Updated dependencies [02ffdcb]
• Updated dependencies [76cb673]
  • @​ai-sdk/provider-utils@​5.0.12

4.0.16

Patch Changes

• afcf19c: fix(provider/anthropic): preserve web search citations when replaying assistant messages
• Updated dependencies [cd06458]
  • @​ai-sdk/provider-utils@​5.0.11

4.0.15

Patch Changes

• Updated dependencies [31c7be8]
  • @​ai-sdk/provider-utils@​5.0.10

... (truncated)

Commits

• c8baafc Version Packages (#17926)
• e29788d fix(anthropic): report thinking tokens as reasoning token usage (#17923)
• 1d44024 Version Packages (#17895)
• cbdc990 feat(anthropic): opus 5 (#17904)
• 3c98985 Version Packages (#17868)
• 01a596a fix: incorrect capability defaults for unrecognized Claude model IDs (#17830)
• 4742a2b Version Packages (#17615)
• 97de198 fix: surface the implicit output token limit for unknown Anthropic models (#1...
• 0319ff3 Version Packages (#17553)
• b72fc7c fix: prevent Bedrock native structured output from forwarding unsupported JSO...
• Additional commits viewable in compare view