@ai-sdk/google
4.0.122 · 17 Nov 2026Apache-2.014.6M / weekRepositoryHomepagenpmAtom feed
497 stable releases, newest first, by release line. Notes are quoted as the maintainers published them.
- 3.x
3.0.7 – 3.0.126 releases with no change for you4 dependency updates, 2 maintenance@ai-sdk/provider 3.0.4@ai-sdk/provider-utils 4.0.8Show each
- 3.0.1222 Jan 2026docs, tests or tooling
- 3.0.1122 Jan 2026docs, tests or tooling
- 3.0.1016 Jan 2026@ai-sdk/provider 3.0.4@ai-sdk/provider-utils 4.0.8
- 3.0.915 Jan 2026@ai-sdk/provider-utils 4.0.7
- 3.0.814 Jan 2026@ai-sdk/provider-utils 4.0.6@ai-sdk/provider 3.0.3
- 3.0.712 Jan 2026@ai-sdk/provider-utils 4.0.5
3.0.6
7 Jan 2026google: Parse structured output when using google provider tools
3.0.5
6 Jan 2026Add file support for Gemini 3 models
3.0.3 – 3.0.42 releases that only update dependencies@ai-sdk/provider 3.0.2@ai-sdk/provider-utils 4.0.4Show each
3.0.2
30 Dec 2025Trigger release to update
@latest
Patch Changes
Trigger release to update
@latest
3.0.1
23 Dec 2025Only updates a dependency:@ai-sdk/provider-utils 4.0.1
3.0.0
22 Dec 2025- Breaking
Ai SDK 6 beta
provider: Add specificationVersion to ProviderV3
provider: Shared spec v3
Major Changes
- Breaking
Ai SDK 6 beta
Patch Changes
provider: Add specificationVersion to ProviderV3
provider: Shared spec v3
google:
thinking_leveloption for Gemini 3provider/google: Preserve nested empty object schemas and descriptions in tool parameters
feat(openai); update spec for mcp approval
The mediaResolution option has been added and is now passed to the Google API.
Support latest gemini model id
Add Google Maps grounding tool support for location-aware Gemini responses
Fixed Zod validation error when using
google.tools.fileSearch(). The Google File Search API returnsfileSearchStoreinstead ofuriinretrievedContext. UpdatedextractSources()function to handle both the old format (Google Search withuri) and new format (File Search withfileSearchStore), maintaining backward compatibility while preventing validation errors. Also fixed title handling to useundefinedfor URL sources and'Unknown Document'for document sources.google: Use dynamic providerOptionsName when retrieving thoughtSignature in convertToGoogleGenerativeAIMessages
When using @ai-sdk/google-vertex provider with Gemini thinking models, multi-step tool calls would fail with "function call is missing a thought_signature" error. This was because thoughtSignature was stored under providerOptions.vertex but retrieved using hardcoded providerOptions.google. This fix passes providerOptionsName to convertToGoogleGenerativeAIMessages and uses it dynamically.
openai: Allow custom names for provider-defined tools
Add Google Vertex RAG Engine grounding provider tool
EmbeddingModelV3Fix adding google search along with url context in vertex ai
provider/google: Remove includethoughts warning
Change streamText loop to merge file part processing into main parts loop
provider/google: Add support for the imageSize provider option
Support for custom provider name in google and anthropic providers
Tool execution approval
Provider-V3Extended token usage
Add provider version to user-agent header
Improve error message when mixing function tools with provider-defined tools to clarify fallback behavior and list ignored function tools
LanguageModelV3provider/google: Lazy schema loading
import time improved by 12.5% (22.3ms ➡️ 19.5ms)
provider/google: Add enterpriseWebSearch tool
google: Add thought signature to gemini 3 pro image parts
Image editing
provider/google: Preserve thoughtSignature through tool execution
vertex: Allow 'vertex' as a key for providerOptions
Add promptFeedback outputs
Prepare search tool for gemini-3-pro-preview
Add Google File search tool
Add
gemini-3-pro-previewandgemini-3-pro-image-previewmodel IDsvertex: Support 'trafficType' in provider usageMetadata
Support
imageConfig.aspectRatioconfiguration for Gemini modelsAdd gemini 3 pro
provider/google: Preserve nested empty object schemas in tool parameters to fix "property is not defined" validation errors when using required properties with empty object types
ImageModelV3Update gemini 3 model id
google: Return request as object
Flexible tool output content support
Expose raw finish reason
google: gemini-3-flash
Moved dependency
@ai-sdk/test-serverto devDependencies@ai-sdk/google: Make title field optional in grounding metadata schema
google,google-vertex: Update known model IDs
Chore (provider/google): Add preview modelIds for gemini 2.5 flash and lite
- 2.x
2.0.111
11 Nov 2026Add support for the
gemini-4-carbonmodel with the same capabilities asgemini-4-argon.
2.0.110
10 Nov 2026Only updates a dependency:@ai-sdk/provider-utils 3.0.46
2.0.109
7 Nov 2026Omit Google provider-defined tools, including File Search and Vertex RAG Store, when
toolChoiceisnoneso disabled tools cannot retrieve data or execute server-side. Preserve function declarations needed by conversation history while disabling function calling.
Patch Changes
Omit Google provider-defined tools, including File Search and Vertex RAG Store, when
toolChoiceisnoneso disabled tools cannot retrieve data or execute server-side. Preserve function declarations needed by conversation history while disabling function calling.
2.0.108
3 Nov 2026google: Handle cycles and BigInt in functionResponse and parse string functionCall arguments
Map Google RPC status names to HTTP status codes when streamed errors omit a numeric code, so transient failures such as
UNAVAILABLEandRESOURCE_EXHAUSTEDare marked retryable.
Patch Changes
google: Handle cycles and BigInt in functionResponse and parse string functionCall arguments
Map Google RPC status names to HTTP status codes when streamed errors omit a numeric code, so transient failures such as
UNAVAILABLEandRESOURCE_EXHAUSTEDare marked retryable.
2.0.107
30 Oct 2026Only updates a dependency:@ai-sdk/provider-utils 3.0.43
2.0.106
23 Oct 2026google: Convert
type: [X, "null"]to a plain nullable node soitemsandformatbeside it stay valid for Gemini
2.0.105
21 Oct 2026Defer validation schema construction until first use to reduce provider import time.
Patch Changes
Defer validation schema construction until first use to reduce provider import time.
2.0.104
19 Oct 2026Preserve grounding metadata from each automatically continued Google response in
groundingMetadataPerRequest. Previously, a later response without grounding could leave earlier Search queries absent from the final provider metadata even though their sources were returned. The new array preserves request order and includesnullfor requests without grounding.
2.0.103
13 Oct 2026google: 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
2.0.102
7 Oct 2026google: Add the 'gemini-nano-banana-2.1' model ID
2.0.101
30 Sep 2026Keep default Node.js downloads protected by DNS validation and connection pinning when frameworks or instrumentation wrap global fetch before or after the SDK loads.
Patch Changes
Keep default Node.js downloads protected by DNS validation and connection pinning when frameworks or instrumentation wrap global fetch before or after the SDK loads.
2.0.99 – 2.0.1002 releases that only update dependencies@ai-sdk/provider-utils 3.0.40Show each
- 2.0.10028 Sep 2026@ai-sdk/provider-utils 3.0.40
- 2.0.9923 Sep 2026@ai-sdk/provider-utils 3.0.39
2.0.98
23 Sep 2026provider: Preserve opaque file URI strings for provider serialization
Patch Changes
provider: Preserve opaque file URI strings for provider serialization
2.0.97
9 Sep 2026Only updates dependencies:@ai-sdk/provider 2.0.4@ai-sdk/provider-utils 3.0.37
2.0.96
3 Sep 2026google: Add gemini-3.8-flash model
2.0.95
31 Aug 2026Only updates a dependency:@ai-sdk/provider-utils 3.0.36
2.0.94
27 Aug 2026Surface prompt-level Google safety blocks without candidates as content-filter results with prompt feedback metadata.
Omit unsupported frequency and presence penalties from Gemini 2.5 requests and return warnings instead.
2.0.93
26 Aug 2026google: Convert enum values to the Gemini schema format
2.0.90
24 Aug 2026Only updates a dependency:@ai-sdk/provider-utils 3.0.33
2.0.89
20 Aug 2026Inline local JSON Schema references in Google tool and structured-output schemas.
2.0.88
13 Aug 2026google: Add
gemini-3.7-flashmodel
2.0.87
11 Aug 2026Preserve Google API error details in
APICallError.data.
Patch Changes
Preserve Google API error details in
APICallError.data.
2.0.86
29 Jul 2026Only updates a dependency:@ai-sdk/provider-utils 3.0.31
2.0.85
23 Jul 2026provider/google: Default unknown Gemini model IDs to the newest supported capabilities
2.0.84
21 Jul 20262.0.83
20 Jul 2026provider/google: Associate multiple code execution results with their tool call
2.0.82
17 Jul 2026Only updates a dependency:@ai-sdk/provider-utils 3.0.30
2.0.81
15 Jul 2026provider/google: Forward Vertex-only imageConfig options (personGeneration, prominentPeople, imageOutputOptions)
2.0.80
14 Jul 2026Only updates a dependency:@ai-sdk/provider-utils 3.0.29
2.0.79
13 Jul 2026Expand standalone Google
thresholdprovider options into safety settings.
2.0.78
1 Jul 2026provider/google: Include usageMetadata in providerMetadata for non-streaming generate
2.0.76 – 2.0.772 releases that only update dependencies@ai-sdk/provider-utils 3.0.28Show each
- 2.0.7729 Jun 2026@ai-sdk/provider-utils 3.0.28
- 2.0.7616 Jun 2026@ai-sdk/provider-utils 3.0.27
2.0.75
12 Jun 2026Only 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, orvideo.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
isSameOriginhelper 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/googlenow 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.
Patch Changes
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, orvideo.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
isSameOriginhelper 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/googlenow 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.
2.0.74
19 May 2026provider/google: Add
gemini-3.5-flash
2.0.73
15 May 2026provider/google: Support
functionCall.idwhen returned by Gemini API and provide matchingfunctionResponse.id
A whole project to check? Paste its package.json in the dependency check: every deprecated, outdated or vulnerable dependency, with what to upgrade to. Free, no account.