Last release today
26 Aug 2026
Ships on a steady schedule
a new release about every 8 days
Some releases are documented
notes for 32 of the last 60 stable releases
Nothing withdrawn
no release was ever pulled
2 years old
588 releases · first in 2024
Release timeline
588 releases since 2024Releases
- 4.0.4426 Aug 2026
- 4.0.4225 Aug 2026
Release notes
Open source →Patch Changes
- 591d25b: feat: add batch completion webhooks.
experimental_startTextBatchaccepts awebhookUrl, and the gateway provider registers it through the batchcallbackUrlcontract and exports typed async-job metadata. Direct Anthropic and OpenAI batch providers return an unsupported warning when the option is provided. - Updated dependencies [591d25b]
- @ai-sdk/[email protected]
- @ai-sdk/[email protected]
- 591d25b: feat: add batch completion webhooks.
- 4.0.4122 Aug 2026
- 4.0.4019 Aug 2026
- 4.0.3914 Aug 2026
Release notes
Open source →Patch Changes
- 4579b08: Preserve Anthropic server-tool caller metadata in multi-turn conversations.
- 4.0.3811 Aug 2026
- 4.0.3710 Aug 2026
- 4.0.367 Aug 2026
Release notes
Open source →Patch Changes
- Updated dependencies [ad6a650]
- Updated dependencies [81cd026]
- @ai-sdk/[email protected]
- @ai-sdk/[email protected]
- 4.0.357 Aug 2026
- 4.0.347 Aug 2026
- 4.0.336 Aug 2026
Release notes
Open source →Patch Changes
- Updated dependencies [3469d0c]
- @ai-sdk/[email protected]
- @ai-sdk/[email protected]
- Updated dependencies [3469d0c]
- 4.0.326 Aug 2026
Release notes
Open source →Patch Changes
- 8b96941: Reject spliced Anthropic generations while allowing duplicate message start events for the active message.
- 4.0.304 Aug 2026
- 4.0.294 Aug 2026
- 4.0.283 Aug 2026
Release notes
Open source →Patch Changes
- 9337ecd: Preserve Anthropic prompt-cache matches by replaying complete code-execution transcripts in their original wire shape.
- Updated dependencies [79e133c]
- @ai-sdk/[email protected]
- @ai-sdk/[email protected]
- 4.0.2731 Jul 2026
Release notes
Open source →Patch Changes
- Updated dependencies [5fc7da5]
- Updated dependencies [93b2acd]
- @ai-sdk/[email protected]
- 4.0.2631 Jul 2026
Release notes
Open source →Patch Changes
- dc0c28e: Return visible summarized reasoning when generic reasoning enables adaptive thinking.
- Updated dependencies [fa95504]
- @ai-sdk/[email protected]
- 4.0.2530 Jul 2026
Release notes
Open source →Patch Changes
- d8210b6: chore: centralize record type guards in provider-utils
- b192878: feat: add experimental_toolCaller routing to generateText for code mode
- Updated dependencies [d8210b6]
- Updated dependencies [b192878]
- @ai-sdk/[email protected]
- 4.0.2429 Jul 2026
Release notes
Open source →Patch Changes
- Updated dependencies [1659cd5]
- Updated dependencies [6a5bdff]
- @ai-sdk/[email protected]
- 4.0.2328 Jul 2026
Release notes
Open source →Patch Changes
- Updated dependencies [0c464d9]
- Updated dependencies [c49380c]
- @ai-sdk/[email protected]
- 4.0.2227 Jul 2026
Release notes
Open source →Patch Changes
- Updated dependencies [1e2f324]
- @ai-sdk/[email protected]
- @ai-sdk/[email protected]
- Updated dependencies [1e2f324]
- 4.0.2126 Jul 2026
Release notes
Open source →Patch Changes
- e29788d: fix(anthropic): report thinking tokens as reasoning token usage
- 4.0.2024 Jul 2026
Release notes
Open source →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)
- 4.0.1923 Jul 2026
Release notes
Open source →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.
- 4.0.1821 Jul 2026
Release notes
Open source →Patch Changes
- 97de198: Warn when an unknown model uses the default 4096 max output token limit.
- 4.0.1721 Jul 2026
Release notes
Open source →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/[email protected]
- 4.0.1617 Jul 2026
Release notes
Open source →Patch Changes
- afcf19c: fix(provider/anthropic): preserve web search citations when replaying assistant messages
- Updated dependencies [cd06458]
- @ai-sdk/[email protected]
- 4.0.1514 Jul 2026
- 4.0.1413 Jul 2026
Release notes
Open source →Patch Changes
-
4be62c1: fix(provider-utils): validate provider-response URLs in
getFromApigetFromApinow has avalidateUrlflag. It is optional so existing callers keep compiling (omitting it behaves likefalse, i.e. no validation), but all AI SDK provider packages set it explicitly at every call site so each one makes a visible trust decision. Whentrue, the URL is routed throughfetchWithValidatedRedirects— the same guard used bydownloadBlob— which rejects private/loopback/link-local targets, re-validates every redirect hop, strips proxy/metadata/cookie request headers, and drops all caller headers except the user-agent on cross-origin redirects (custom API-key headers must not follow a redirect off-origin any more thanAuthorizationmay); blocked URLs throwDownloadError. It is enabled at the image/video/audio download and polling call sites where the URL comes from a provider response body; URLs built from developer-configured endpoints passvalidateUrl: falseand are unaffected.A new optional
credentialedOriginwithholds caller headers unless the URL is same-origin with it, so the API key is not sent to a response-supplied host on a different origin.A new optional
trustedOriginexempts URLs (and redirect hops) that are same-origin with the developer-configured provider endpoint from target validation, so self-hosted and localhost deployments whose response URLs point back at the configured host keep working; all other hops are still validated.Also closes range gaps in
validateDownloadUrl(IPv4224.0.0.0/4multicast and the TEST-NET documentation ranges192.0.2.0/24,198.51.100.0/24,203.0.113.0/24; IPv6 documentation ranges2001:db8::/32and3fff::/20), and follows only the fetch-spec redirect status codes (301/302/303/307/308) — aLocationheader on any other status is not followed. This guard performs string/literal checks only and does not resolve DNS; hostnames that resolve to private addresses and DNS rebinding remain out of scope and must be constrained at the network layer (or by injecting a Nodefetchthat pins the resolved IP at connect time) for server deployments handling untrusted URLs. Seecontributing/secure-url-handling.md. -
cd12954: Reject empty OpenAI, Anthropic, and Replicate base URLs with a helpful AI SDK invalid argument error.
-
Updated dependencies [4be62c1]
-
Updated dependencies [7805e4a]
-
Updated dependencies [cd12954]
- @ai-sdk/[email protected]
-
- 4.0.1210 Jul 2026
Release notes
Open source →Patch Changes
- 308a519: chore: enforce consistent imports from
zod/v4instead ofzod
- 308a519: chore: enforce consistent imports from
- 4.0.119 Jul 2026
Release notes
Open source →Patch Changes
- Updated dependencies [0f93c57]
- @ai-sdk/[email protected]
- @ai-sdk/[email protected]
- Updated dependencies [0f93c57]
- 4.0.108 Jul 2026
- 4.0.97 Jul 2026
- 4.0.84 Jul 2026
Release notes
Open source →Patch Changes
-
0aa0ff3: fix(anthropic): forward
thinking: { type: 'disabled' }to the API instead of stripping itPreviously, setting
providerOptions.anthropic.thinking = { type: 'disabled' }(or top-levelreasoning: 'none') was accepted by the schema but silently dropped from the outgoing request. For models that default thinking on (e.g. Sonnet 5), this left thinking enabled and could consume a smallmax_tokensbudget entirely. Thedisabledvalue is now sent to the Anthropic Messages API.
-
- 4.0.72 Jul 2026
Release notes
Open source →Patch Changes
- 5c5c0f5: Add experimental streaming transcription support for transcription models, including OpenAI
gpt-realtime-whisperand xAI WebSocket STT. - Updated dependencies [5c5c0f5]
- @ai-sdk/[email protected]
- @ai-sdk/[email protected]
- 5c5c0f5: Add experimental streaming transcription support for transcription models, including OpenAI
- 4.0.62 Jul 2026
Release notes
Open source →Patch Changes
- c6f5e62: Prevent prototype pollution when synchronously parsing provider JSON inputs and expose
secureJsonParsefrom provider-utils. - 679c52a: Normalize a bare
https://api.anthropic.combase URL to include/v1. - Updated dependencies [c6f5e62]
- @ai-sdk/[email protected]
- c6f5e62: Prevent prototype pollution when synchronously parsing provider JSON inputs and expose
- 4.0.51 Jul 2026
- 4.0.430 Jun 2026
Release notes
Open source →Patch Changes
- c18018c: feat (provider/anthropic): add
claude-sonnet-5model id
- c18018c: feat (provider/anthropic): add
- 4.0.330 Jun 2026
Release notes
Open source →Patch Changes
- Updated dependencies [0274f34]
- @ai-sdk/[email protected]
- @ai-sdk/[email protected]
- Updated dependencies [0274f34]
- 4.0.229 Jun 2026
Release notes
Open source →Patch Changes
- dfffb27: fix(anthropic): correctly map delta type when code execution tools are used
- 4.0.127 Jun 2026
- 4.0.025 Jun 2026
Release notes
Open source →Major Changes
-
832f86f: fix(anthropic): remove cacheCreationInputTokens from providerMetadata
-
34bd95d: feat(ai): add support for uploading provider skills using the provider references abstraction
-
ef992f8: Remove CommonJS exports from all packages. All packages are now ESM-only (
"type": "module"). Consumers usingrequire()must switch to ESMimportsyntax. -
c29a26f: feat(provider): add support for provider references and uploading files as supported per provider
-
3887c70: feat(provider): add new top-level reasoning parameter to spec and support it in
generateTextandstreamText -
8359612: Start v7 pre-release
-
04e9009: chore: make provider implementations code patterns more consistent, including renaming certain exported symbols
For all externally exported symbols that were renamed, the old names continue to work via deprecated aliases.
Patch Changes
-
e02f041: feat(provider/anthropic): add support for
claude-opus-4-8 -
648705c: fix(provider/anthropic): fix remaining errors with Anthropic
code_executiontool dynamic calls from latestweb_fetchorweb_search -
38fc777: Add AI Gateway hint to provider READMEs
-
ee798eb: chore(provider-utils): rename
Experimental_SandboxtoExperimental_SandboxSession -
e748b35: chore: update v3 specs to v4
-
19c5ee2: fix(anthropic): reorder assistant content b/w client and provider tool use
-
0ee8aec: feat (provider/anthropic): support passing metadata.user_id
-
c012d57: feat(anthropic): sanitize the unsupported JSON schema validation properties
-
8018480: feat(anthropic): add the new advisor tool
-
e5c4f40: Remove stale
effort-2025-11-24beta header — the extended thinking effort parameter is GA and no longer requires the beta flag. Vertex AI's strict validator was actively rejecting requests with this header. -
f57c702: fix(anthropic): allow both temperature and topP for non-Anthropic models using the Anthropic-compatible API
The temperature/topP mutual exclusivity check now only applies to known Anthropic models (model IDs starting with
claude-). Non-Anthropic models using the Anthropic-compatible API (e.g. Minimax) can now send both parameters as required by their APIs. -
2610e84: feat(provider/anthropic): automatically use sandbox in bash tool
-
d848405: feat: add optional
abortSignalparameters to sandbox command execution -
87d1723: chore(anthropic): remove unnecessary messages affix from Anthropic symbols and files
-
9f0e36c: trigger release for all packages after provenance setup
-
acdbf84: Handle errors from anthropic websearch tool
-
ad0b376: fix(provider/anthropic): stop adding
fine-grained-tool-streaming-2025-05-14beta forclaude-opus-4-7 -
58a2ad7: fix: more precise default message for tool execution denial
-
21d1ee3: fix(anthropic): skip passing beta header for tool search tools
-
1fe058b: fix(anthropic): preserve the error code returned by model
-
7fc6bd6: Raise minimum supported Node.js version to 22. Supported versions: 22, 24, and 26.
-
09bd27b: feat (provider/anthropic): add support for inference_geo provider option
-
0c4c275: trigger initial canary release
-
6fd51c0: fix(provider): preserve error type prefix in getErrorMessage
-
e311194: feat(ai): allow passing provider instance to
uploadFileanduploadSkillas shorthand -
6c93e36: feat(provider-utils): add
spawnCommandmethod toExperimental_Sandboxto allow for detached command execution -
9bd6512: feat(provider): change file part data property to be tagged with a type and remove the image part type
-
258c093: chore: ensure consistent import handling and avoid import duplicates or cycles
-
5463d0d: feat(provider): align tool result output content file part types with top-level message file part types
-
b8396f0: trigger initial beta release
-
6b4d325: feat(provider/anthropic): add support for
claude-fable-5and thefallbacksAPI parameter -
f05a40d: fix(vertex): throw warning when strict: true for vertexAnthropic
-
a6617c5: feat(provider-utils): add
readFileandwriteFileplus convenience wrappers toExperimental_Sandboxabstraction -
a464505: fix(anthropic): propagate toModelOutput providerOption to anthropic tool results
-
90e2d8a: chore: fix unused vars not being flagged by our lint tooling
-
b3976a2: Add workflow serialization support to all provider models.
@ai-sdk/provider-utils: NewserializeModel()helper that extracts only serializable properties from a model instance, filtering out functions and objects containing functions. Third-party provider authors can use this to add workflow support to their own models.All providers:
headersis now optional in provider config types. This is non-breaking — existing code that passesheaderscontinues to work. Custom provider implementations that construct model configs manually can now omitheaders, which is useful when models are deserialized from a workflow step boundary where auth is provided separately.All provider model classes now include
WORKFLOW_SERIALIZEandWORKFLOW_DESERIALIZEstatic methods, enabling them to cross workflow step boundaries without serialization errors. -
0d8f107: feat(provider/anthropic): add support for Opus 4.7 and relevant API enhancements
-
ff5eba1: feat: roll
image-*tool output types into their equivalentfile-*types
-
- 4.0.0-canary.6612 Jun 2026pre-release
Release notes
Open source →Patch Changes
- Updated dependencies [aeda373]
- Updated dependencies [375fdd7]
- Updated dependencies [b4507d5]
- @ai-sdk/[email protected]
- 4.0.0-canary.6511 Jun 2026pre-release
- 4.0.0-canary.649 Jun 2026pre-release
Release notes
Open source →Patch Changes
- 6b4d325: feat(provider/anthropic): add support for
claude-fable-5and thefallbacksAPI parameter
- 6b4d325: feat(provider/anthropic): add support for
- 4.0.0-canary.635 Jun 2026pre-release
Release notes
Open source →Patch Changes
- Updated dependencies [ce769dd]
- @ai-sdk/[email protected]
- @ai-sdk/[email protected]
- Updated dependencies [ce769dd]
- 4.0.0-canary.624 Jun 2026pre-release
Release notes
Open source →Patch Changes
- ee798eb: chore(provider-utils): rename
Experimental_SandboxtoExperimental_SandboxSession - Updated dependencies [ee798eb]
- Updated dependencies [daf6637]
- @ai-sdk/[email protected]
- ee798eb: chore(provider-utils): rename
- 4.0.0-canary.6128 May 2026pre-release
Release notes
Open source →Patch Changes
- e02f041: feat(provider/anthropic): add support for
claude-opus-4-8
- e02f041: feat(provider/anthropic): add support for
- 4.0.0-canary.6026 May 2026pre-release
Release notes
Open source →Patch Changes
- 6c93e36: feat(provider-utils): add
spawnCommandmethod toExperimental_Sandboxto allow for detached command execution - Updated dependencies [6c93e36]
- Updated dependencies [f617ac2]
- @ai-sdk/[email protected]
- 6c93e36: feat(provider-utils): add
- 4.0.0-canary.5922 May 2026pre-release
Release notes
Open source →Patch Changes
- 648705c: fix(provider/anthropic): fix remaining errors with Anthropic
code_executiontool dynamic calls from latestweb_fetchorweb_search
- 648705c: fix(provider/anthropic): fix remaining errors with Anthropic
- 4.0.0-canary.5822 May 2026pre-release
- 4.0.0-canary.5719 May 2026pre-release
Release notes
Open source →Patch Changes
- 7fc6bd6: Raise minimum supported Node.js version to 22. Supported versions: 22, 24, and 26.
- Updated dependencies [7fc6bd6]
- @ai-sdk/[email protected]
- @ai-sdk/[email protected]
- 4.0.0-canary.5515 May 2026pre-release
Release notes
Open source →Patch Changes
- a464505: fix(anthropic): propagate toModelOutput providerOption to anthropic tool results
- Updated dependencies [28dfa06]
- Updated dependencies [e93fa91]
- @ai-sdk/[email protected]
- 4.0.0-canary.5413 May 2026pre-release
- 4.0.0-canary.5313 May 2026pre-release
- 4.0.0-canary.5212 May 2026pre-release
- 4.0.0-canary.5111 May 2026pre-release
- 4.0.0-canary.508 May 2026pre-release
Release notes
Open source →Patch Changes
- d848405: feat: add optional
abortSignalparameters to sandbox command execution - Updated dependencies [d848405]
- @ai-sdk/[email protected]
- d848405: feat: add optional
- 4.0.0-canary.498 May 2026pre-release
- 4.0.0-canary.487 May 2026pre-release