NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
npm, #4234 most downloaded on npm
Latest stable
4.0.117
Last release 7 days ago
10 Nov 2026
Ships on a steady schedule
a new release about every 8 days
Most releases are documented
notes for 42 of the last 60 stable releases
Nothing withdrawn
no release was ever pulled
3 years old
743 releases · first in 2024
Drag across the ruler to compare, or tap a release to open it.
a4ba394 : feat: support per-request models in batch
048ce06 : feat(batch): surface the uploaded input file on the batch start result ( providerMetadata.<provider>.inputFileId / inputFileExpiresAt ) and
Your coding agent can read these notes before it upgrades. Set up the MCP server →
providerMetadata.<provider>.inputFileId / inputFileExpiresAt) and accept an inputFileExpiresAfter provider option on the OpenAI and xAI batch input file uploadgemini-3.5-transcribe) via generateContent with language detection, speaker diarization, word timestamps, and custom vocabulary, plus streaming transcription (gemini-3.5-transcribe-live) over the Live API WebSocket with mode: 'VERBATIM' | 'SMART' transcription formattingUpdated dependencies [ 6bcc0f8 ]
e7fc90e: feat(google): support the Gemini Batch API with experimental_startTextBatch
### Patch Changes - Updated dependencies [b74971f] - @ai-sdk/provider-utils@5.0.29
Updated dependencies [ 90192f1 ]
6c5a1ed: Inline local JSON Schema references in Google tool and structured-output schemas.
Updated dependencies [ 3e125ba ]
f69920a: fix(google): use low as the minimum reasoning level for full Gemini Flash 3.7 and later
bb0cf2e: fix(google): coerce minimal reasoning to low for Gemini 3.7 Flash
16650e9: feat(google): add gemini-3.7-flash model
gemini-3.7-flash model### Patch Changes - Updated dependencies [b6fff2e] - @ai-sdk/openai@4.0.42
a062795: fix(openai): support built-in and provider-defined tools in the Responses allowedTools option
allowedTools emitted every allow-list entry as { type: 'function', name }, but OpenAI identifies
built-in tools by type. Allow-listing a declared provider-defined tool (web search, image generation,
MCP, custom, ...) therefore failed with Tool choice '<name>' not found in 'tools' parameter. Entries
are now derived from the declared tool, including the MCP server label and custom tool name.
Tools that OpenAI cannot allow-list (the tool search tool, deferred tools, and namespaced tools) are dropped from the allow-list with a warning, and an error is thrown if that would leave the allow-list empty rather than silently sending an unrestricted request.
Ambiguous names are now reported instead of resolved silently. A name that matches both a declared tool and another tool's provider tool name resolves to the declared tool and warns; a provider tool name shared by several tools in the same request (two MCP servers, for example) is dropped with a warning. A name that matches no declared tool keeps its existing behavior and is now warned about.
b6fff2e: feat(provider/openai): support explicit Responses compaction triggers
da78d58: fix(google): convert enum values to the Gemini schema format
APICallError.data.1ffa1d2: feat(xai): speech timestamps, pronunciation replacements, provider metadata, and error parsing
1ffa1d2: feat(xai): speech timestamps, pronunciation replacements, provider metadata, and error parsing
withTimestamps and replace provider options for text to speech. With
withTimestamps, the JSON envelope is decoded and the audio returned as
usual, while duration, content type, and character-level alignment are
exposed via providerMetadata.xai.providerMetadata.xai.traceId (from the x-trace-id response
header) on every speech response.{"error":"..."}) so APICallError
messages carry xAI's real error detail instead of the HTTP reason phrase.4579b08: Preserve Anthropic server-tool caller metadata in multi-turn conversations.
484293f: feat(xai): support the priority service tier on chat and responses
a4d386d: feat(xai): add the Grok 4.6 model IDs and support its xhigh reasoning effort
xhigh reasoning effort*TranslationModel and its related types to *SpeechTranslationModel for consistency### Patch Changes - Updated dependencies [3469d0c] - @ai-sdk/provider@4.0.6 - @ai-sdk/provider-utils@5.0.23
### Patch Changes - Updated dependencies [2b60826] - @ai-sdk/provider-utils@5.0.22
### Patch Changes - Updated dependencies [1bec07d] - @ai-sdk/provider-utils@5.0.21
### Patch Changes - Updated dependencies [160ccdb] - @ai-sdk/provider-utils@5.0.20
79e133c: async APIs for generateVideo (poll, webhook)
79e133c: async APIs for generateVideo (poll, webhook)
Adds an asynchronous start/status flow to the experimental video model
interface (VideoModelV4): models may now implement doStart, doStatus,
and handleWebhookOption instead of (or in addition to) doGenerate, and
experimental_generateVideo accepts poll and webhook options to
orchestrate completion via polling or webhooks. Polling configuration can use
a custom delay implementation for durable workflow compatibility.
Updated dependencies [79e133c]
5fc7da5: chore: centralize empty language model usage creation in provider utilities.
### Patch Changes - Updated dependencies [fa95504] - @ai-sdk/provider-utils@5.0.17
d8210b6: chore: centralize record type guards in provider-utils
### Patch Changes - Updated dependencies [1659cd5] - Updated dependencies [6a5bdff] - @ai-sdk/provider-utils@5.0.15
d2d9324: Forward topK through Google Interactions requests and warn when unsupported frequency or presence penalties are provided.
topK through Google Interactions requests and warn when unsupported frequency or presence penalties are provided.c49380c: feat: add experimental streaming speech translation models (openai.translation('gpt-realtime-translate') over the OpenAI Realtime translation
openai.translation('gpt-realtime-translate') over the OpenAI Realtime translations WebSocket and google.translation('gemini-3.5-live-translate-preview') over the Gemini Live API). connectToWebSocket in @ai-sdk/provider-utils now passes close code and reason to onClose (additive, optional parameter).### Patch Changes - Updated dependencies [1e2f324] - @ai-sdk/provider@4.0.4 - @ai-sdk/provider-utils@5.0.13
c57a353: fix(google): omit unsupported function call IDs
f126649: feat(provider/google): default unknown Gemini model IDs to the newest supported capabilities
66b7151: fix(provider/google): preserve Gemini Live lifecycle events
7c16f21: feat(google): add gemini-3.6-flash and gemini-3.5-flash-lite models
gemini-3.6-flash and gemini-3.5-flash-lite models### Patch Changes - Updated dependencies [02ffdcb] - Updated dependencies [76cb673] - @ai-sdk/provider-utils@5.0.12
a70b027: fix(provider/google): associate multiple code execution results with their tool call
responseId as response-metadata (stream) and response.id (generate)### Patch Changes - Updated dependencies [cd06458] - @ai-sdk/provider-utils@5.0.11
5b4a299: fix(provider/google): forward Vertex-only imageConfig options (personGeneration, prominentPeople, imageOutputOptions)
662ddfc: Allow google.interactions agent requests to include supported tools, including file_search.
### Patch Changes - Updated dependencies [31c7be8] - @ai-sdk/provider-utils@5.0.10
4be62c1: fix(provider-utils): validate provider-response URLs in getFromApi
4be62c1: fix(provider-utils): validate provider-response URLs in getFromApi
getFromApi now has a validateUrl flag. It is optional so existing callers keep compiling (omitting it behaves like false, 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. When true, the URL is routed through fetchWithValidatedRedirects — the same guard used by downloadBlob — 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 than Authorization may); blocked URLs throw DownloadError. 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 pass validateUrl: false and are unaffected.
A new optional credentialedOrigin withholds 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 trustedOrigin exempts 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 (IPv4 224.0.0.0/4 multicast and the TEST-NET documentation ranges 192.0.2.0/24, 198.51.100.0/24, 203.0.113.0/24; IPv6 documentation ranges 2001:db8::/32 and 3fff::/20), and follows only the fetch-spec redirect status codes (301/302/303/307/308) — a Location header 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 Node fetch that pins the resolved IP at connect time) for server deployments handling untrusted URLs. See contributing/secure-url-handling.md.
Updated dependencies [4be62c1]
Updated dependencies [7805e4a]
Updated dependencies [cd12954]
e40118c: Forward text file references as documents when using the Google Interactions API.
17d66c5: Fix Google tool result conversion to send file data as inline data instead of JSON text on the legacy tool-result path.
threshold provider options into safety settings.inputReferences for reference-to-video generation### Patch Changes - Updated dependencies [ac306ed] - @ai-sdk/provider-utils@5.0.6
7401c2c: Pass documented Gemini external HTTPS file URLs through without downloading them.
5c5c0f5: Add experimental streaming transcription support for transcription models, including OpenAI gpt-realtime-whisper and xAI WebSocket STT.
gpt-realtime-whisper and xAI WebSocket STT.c6f5e62: Prevent prototype pollution when synchronously parsing provider JSON inputs and expose secureJsonParse from provider-utils.
secureJsonParse from provider-utils.d20f0dc: Add vertex.interactions() for the Gemini Interactions API on Vertex AI. Targets the location-scoped .../locations/{region}/interactions resou
vertex.interactions() for the Gemini Interactions API on Vertex AI. Targets the location-scoped .../locations/{region}/interactions resource using the existing Vertex OAuth credentials, enabling multimodal-output models such as gemini-omni-flash-preview (video output) through Vertex. The GoogleInteractionsLanguageModel is now exported from @ai-sdk/google/internal for provider reuse.### Patch Changes - Updated dependencies [8c616f0] - @ai-sdk/provider-utils@5.0.3
dc1eb8d: Support Gemini Interactions video output. Parse video output blocks from the Google Interactions API into file parts (buffered and streaming)
providerMetadata.google.outputTokensByModality.0274f34: feat (video): add first-class frameImages and inputReferences call options for video generation
frameImages and inputReferences call options for video generation### Patch Changes - Updated dependencies [6a436e3] - @ai-sdk/provider-utils@5.0.1
ba6d510: chore: fix deprecated use of zod .passthrough()
.passthrough()4e825f3: feat(google): update Interactions API implementation to cater for upstream breaking changes coming May 26
a3757d7: chore(provider/google): remove useless GenerativeAI affix from types and classes for consistency
f7d4f01: feat(provider): add support for reasoning-file type for files that are part of reasoning
ef992f8: Remove CommonJS exports from all packages. All packages are now ESM-only ("type": "module"). Consumers using require() must switch to ESM import syntax.
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 generateText and streamText
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.
7f04802: feat(provider/google): add support for managed agents in the Interactions API
38fc777: Add AI Gateway hint to provider READMEs
b71c0d7: feat(provider/google): support Google search grounding when using generateImage with Gemini
6190649: chore(provider/google): remove obsolete Google image model
55db546: fix(provider/google): fix Gemini service tier enum after upstream update
9a0a618: fix(google): fix serviceTier to be correctly formatted for Vertex API
00a0f36: feat(google, google-vertex): added gemini-embedding-2, deep-research-max-preview-04-2026 and deep-research-preview-04-2026
9715ec7: feat(provider/google): add support for service tier parameter
41da50c: fix(provider/google): support functionCall.id when returned by Gemini API and provide matching functionResponse.id
add4326: fix(provider/google): correct JSDoc for multimodal embedding content option
a8d70b6: fix(google): auto-inject skip_thought_signature_validator for Gemini 3 tool-call replays without a signature
947cdab: Add Google realtime Live Translation provider options for gemini-3.5-live-translate-preview.
4e825f3: feat(google): update Interactions API implementation to cater for upstream breaking changes coming May 26
b563707: feat(provider/google): add Gemini text-to-speech (speech) model support
a05109d: feat(provider/google): preserve per-modality token details in usage data
cfca634: fix(google): emit Vertex no-args streaming tool calls and preserve thoughtSignature
Vertex emits a no-args function call as a single chunk shaped { functionCall: { name: 'X' } } with no args, no partialArgs, and no willContinue. The streaming parser had no branch for this shape, so the call was dropped along with any thoughtSignature it carried. For Gemini 3 thinking models this caused the next multi-turn step to 400 with missing thought_signature. The unary (doGenerate) path had the same drop.
Both paths now emit the call as a complete tool call with '{}' input and propagate thoughtSignature provider metadata.
Fixes #14847.
c9c4661: fix(provider/google): preserve groundingMetadata and urlContextMetadata when they arrive in a stream chunk before the finishReason chunk
045d2e8: fix(google): read serviceTier from usageMetadata.serviceTier in both generate and stream paths
The previous implementation read serviceTier from the x-gemini-service-tier
response header, which is only populated on non-streaming responses. Gemini
streaming includes the value in usageMetadata.serviceTier on every chunk, so
providerMetadata.google.serviceTier was always null for streams. Read from
usageMetadata for both paths instead.
19b95f9: fix(google): use VALIDATED function calling mode when any tool has strict:true
2ce3c65: feat(provider/google-vertex): add Gemini text-to-speech (speech) model support
5878b40: fix(google): prevent prototype pollution when streaming tool args
46d1149: chore(provider-utils,google): fix grammar errors in error and warning messages
6a26901: feat(google): add fileData support to embedding model
01fa606: feat(provider/google): support combining built-in tools with function calling on Gemini 3
5036db8: feat(google-vertex): add support for streaming tool arguments input
9f0e36c: trigger release for all packages after provenance setup
a2609df: fix(google): fix streaming tool call args
ab43029: feat(provider/google): support multimodal content parts in embedding provider options
82288b0: feat(provider/google): add gemini-embedding-2-preview and fix multimodal embedding support with embedMany
b04e23e: feat(provider/google): add support for the Gemini Interactions API
58a2ad7: fix: more precise default message for tool execution denial
e59176c: chore(google): update v3 specs to v4
ebbb0f2: fix(provider/google): fix lack of image consistency when using Interactions API in stateless mode
aeda373: fix: only send provider credentials to same-origin response-supplied URLs
Several provider clients followed a URL taken from the provider's API response (a polling/status URL or a final media URL such as polling_url, urls.get, result_url, result.sample, or video.uri) and reused the authenticated headers — or appended ?key=<API_KEY> — on that request. Because the host of the response-supplied URL was never validated, the long-lived API key was sent to whatever host the response named (a CDN in the benign case, or an attacker-chosen host if the provider response was tampered with), allowing credential exfiltration.
A new isSameOrigin helper is added to @ai-sdk/provider-utils, and the affected fetches in @ai-sdk/black-forest-labs, @ai-sdk/fireworks, @ai-sdk/replicate, @ai-sdk/gladia, @ai-sdk/fal, and @ai-sdk/google now attach credentials only when the followed URL is same-origin with the provider's configured API origin. Requests to a foreign origin are made without the credential.
7fc6bd6: Raise minimum supported Node.js version to 22. Supported versions: 22, 24, and 26.
0c4c275: trigger initial canary release
bdb9ea1: chore(google): remove now obsolete Interactions API Api-Revision header
db394ab: feat(provider/google): support cancelling long-running Interactions API agents via AbortSignal, and process their intermittent stream
ce769dd: feat(provider): add experimental Realtime API support for voice conversations
Adds first-class support for realtime (speech-to-speech) APIs:
Experimental_RealtimeModelV4 spec in @ai-sdk/provider with normalized event types and factoryopenai.experimental_realtime() / google.experimental_realtime() / xai.experimental_realtime() work in both server and browser.getToken() static method on each provider for server-side ephemeral token creationexperimental_getRealtimeToolDefinitions helper for provider session tool definitionsexperimental_useRealtime hook in @ai-sdk/react returning UIMessage[] (aligned with useChat), with onToolCall and addToolOutput for client-driven tool executioninputAudioTranscription session config for showing transcribed user audio messages when supported by the provider5b7e7c2: fix(google-vertex): don't send streamFunctionCallArguments for unary API calls and change default to false
2edd14e: fix(provider/google): correctly mark reasoning files as such and fix related multi-turn errors
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
546cefe: feat(provider/google): add gemini-3.5-flash
5463d0d: feat(provider): align tool result output content file part types with top-level message file part types
55f0938: chore(provider/google): update available models
b8396f0: trigger initial beta release
84f36e0: fix(google): omit passing includeServerSideToolInvocations for Vertex tool_config
90e2d8a: chore: fix unused vars not being flagged by our lint tooling
4dac142: feat(google): add new finishMessage field in providerMetadata
aeea161: feat(google): read serviceTier from x-gemini-service-tier response header in Gemini API and use PayGo for Vertex
18c1970: feat(provider/google): Add multimodal tool-result support for Google function responses.
Tool results with output.type = 'content' now map media parts into
functionResponse.parts for Google models, including image-data,
file-data, and base64 data: URLs in URL-style content parts.
Remote HTTP(S) URLs in URL-style tool-result parts are not supported.
b3976a2: Add workflow serialization support to all provider models.
@ai-sdk/provider-utils: New serializeModel() 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: headers is now optional in provider config types. This is non-breaking — existing code that passes headers continues to work. Custom provider implementations that construct model configs manually can now omit headers, which is useful when models are deserialized from a workflow step boundary where auth is provided separately.
All provider model classes now include WORKFLOW_SERIALIZE and WORKFLOW_DESERIALIZE static methods, enabling them to cross workflow step boundaries without serialization errors.
ff5eba1: feat: roll image-* tool output types into their equivalent file-* types
5878b40: fix(google): prevent prototype pollution when streaming tool args
5878b40: fix(google): prevent prototype pollution when streaming tool args
aeda373: fix: only send provider credentials to same-origin response-supplied URLs
Several provider clients followed a URL taken from the provider's API response (a polling/status URL or a final media URL such as polling_url, urls.get, result_url, result.sample, or video.uri) and reused the authenticated headers — or appended ?key=<API_KEY> — on that request. Because the host of the response-supplied URL was never validated, the long-lived API key was sent to whatever host the response named (a CDN in the benign case, or an attacker-chosen host if the provider response was tampered with), allowing credential exfiltration.
A new isSameOrigin helper is added to @ai-sdk/provider-utils, and the affected fetches in @ai-sdk/black-forest-labs, @ai-sdk/fireworks, @ai-sdk/replicate, @ai-sdk/gladia, @ai-sdk/fal, and @ai-sdk/google now attach credentials only when the followed URL is same-origin with the provider's configured API origin. Requests to a foreign origin are made without the credential.
Updated dependencies [aeda373]
Updated dependencies [375fdd7]
Updated dependencies [b4507d5]
@ai-sdk/provider-utils@5.0.0-canary.47
ce769dd: feat(provider): add experimental Realtime API support for voice conversations
ce769dd: feat(provider): add experimental Realtime API support for voice conversations
Adds first-class support for realtime (speech-to-speech) APIs:
Experimental_RealtimeModelV4 spec in @ai-sdk/provider with normalized event types and factoryopenai.experimental_realtime() / google.experimental_realtime() / xai.experimental_realtime() work in both server and browser.getToken() static method on each provider for server-side ephemeral token creationexperimental_getRealtimeToolDefinitions helper for provider session tool definitionsexperimental_useRealtime hook in @ai-sdk/react returning UIMessage[] (aligned with useChat), with onToolCall and addToolOutput for client-driven tool executioninputAudioTranscription session config for showing transcribed user audio messages when supported by the providerUpdated dependencies [ce769dd]
2ce3c65: feat(provider/google-vertex): add Gemini text-to-speech (speech) model support
@ai-sdk/provider-utils@5.0.0-canary.45
b563707: feat(provider/google): add Gemini text-to-speech (speech) model support
One column per month.