🤖 I have created a release beep boop

1.50.0 (2026-04-14)

Features

• [Experimental] Add interactions support. (abc1ee4)
• [Python] Add consent_audio and voice_consent_signature and AsyncSession.setup_complete (959c01b)
• [Python] add RegisterFiles so gcs files can be used with genai. (10bfaec)
• [Python] Implement async embedding batches for MLDev. (12ddfce)
• Add 'turn_complete_reason' and 'waiting_for_input' fields. (5bc4873)
• Add "eu" as a supported service location for Vertex AI platform. (78e1bc1)
• Add addWatermark parameter to the edit image configuration. (c4598da)
• Add ImageConfig to GenerateContentConfig (6fb5eba)
• Add images() convenience method to GenerateImagesResponse (155df8d)
• Add NO_IMAGE enum value to FinishReason (6b00c0b)
• Add registerFiles for Java. (ce0b638)
• add sdkHttpResponse.headers to *Delete responses. (4be038d)
• Add VideoGenerationMaskMode enum for Veo 2 Editing (e5c8277)
• Add add_watermark field for recontext_image (Virtual Try-On, Product Recontext) (5aacbc0)
• add an _exclude_fields characteristic in the config for unused values to avoid unnecessary and unused code in converter and type generation. (bee4e3e)
• Add clearXxx methods to data type builders (a4900c9)
• add complete stats to BatchJob (659c65c)
• add display name to FunctionResponseBlob (8db8c57)
• add display name to FunctionResponseFileData (f5ee8b7)
• Add empty response for tunings.cancel() (57218f5)
• Add enable_enhanced_civic_answers in GenerationConfig (684a2c5)
• Add enableEnhancedCivicAnswers feature in GenerateContentConfig (0570478)
• Add evaluation support to Vertex tuning (7c0bb19)
• Add EvaluationConfig support to tune() in Java (795cf73)
• Add FileSearch tool and associated FileSearchStore management APIs (8ada6ef)
• Add FileSearch tool and associated FileSearchStore management APIs (e4df737)
• Add FunctionResponsePart & ToolComputerUse.excludedPredefinedFunctions (1a24bed)
• Add FunctionResponsePart & ToolComputerUse.excludedPredefinedFunctions (99b50a6)
• Add FunctionResponsePart & ToolComputerUse.excludedPredefinedFunctions (ea954d3)
• Add FunctionResponsePart & ToolComputerUse.excludedPredefinedFunctions (cf8b8b6)
• Add FunctionResponsePart & ToolComputerUse.excludedPredefinedFunctions (f0aebc2)
• Add gemini-3-pro-preview support for local tokenizer (40480f4)
• Add generate_content_config.thinking_level (a47df92)
• Add generateVideosFromSource in Go, refactor private generateVideos (c683835)
• Add Image Grounding support to GoogleSearch tool (0daefbc)
• Add image output options to ImageConfig for Vertex (3eac0b8)
• Add IMAGE_RECITATION and IMAGE_OTHER enum values to FinishReason (26c3c35)
• Add image_size field for Gemini Developer API Imagen 4 generation (c50c755)
• Add image_size field for Vertex Imagen 4 generation (950c0c6)
• Add image_size to ImageConfig (Early Access Program) (c1af981)
• Add inference_generation_config to EvaluationConfig for Tuning (c9632f0)
• Add labels field to Imagen configs (e69cf68)
• Add labels field to Veo configs (86f235a)
• Add labels to create tuning job config (695e17a)
• Add Live Avatar new fields (8e8146a)
• Add minimal and medium thinking levels. (ecfadfe)
• Add minimal and medium thinking levels. (220bf5b)
• Add minimal and medium thinking levels. (bc2ed46)
• Add model_status to GenerateContentResponse (Gemini API only) (6c35249)
• Add ModelArmorConfig support for prompt and response sanitization via the Model Armor service (9c77a8f)
• Add new languages for Imagen 4 prompt language (7e1e6d2)
• Add output_gcs_uri to Imagen upscale_image (7649467)
• Add part_metadata in Part (Gemini API only) (6c35249)
• Add part.media_resolution (a47df92)
• Add PersonGeneration to ImageConfig for Vertex Gempix (d8a4c43)
• Add ProxyOptions in ClientOptions for configuring proxies (eafdf79)
• Add resolution field for Gemini Developer API Veo 3 generation (eec410c)
• Add safety_filter_level and person_generation for Imagen upscaling (09a8075)
• Add service tier to GenerateContent. (5f61a16)
• Add support for preference optimization tuning in the SDK. (5d4123c)
• add tests for the async Operations module (14c008a)
• add the response body for generateContent (a011580)
• add the response body for generateContent (6e28ab4)
• add the response body for generateContent (b2a5b3f)
• Add thinking_config for live (274c21d)
• add Tuning support for Java (0ab209d)
• Add ultra high resolution to the media resolution in Parts. (7c9b7f6)
• Add UnifiedMetric support to Vertex Tuning evaluation config (a28ebfc)
• Add utility methods for creating FunctionResponsePart and creating FunctionResponse Part with FunctionResponseParts (af16a4c)
• Add VALIDATED mode into FunctionCallingConfigMode (4bb8680)
• Add VideoGenerationReferenceType enum for generate_videos (df9d910)
• Add voice activity detection signal. (d1ca685)
• Add webhook_config to batches.create() and models.generate_videos() (ac0e49e)
• Added Operations.get which is a generic method which will handle all Operation types. (c1dc32f)
• Added phish filtering feature. (ed4e12c)
• Allow custom endpoints for authentication with Vertex AI in Java (5b38728)
• allow methods in batch to return headers in sdk_http_response by default (f937dd1)
• allow methods in models to return headers in sdk_http_response by default. (7415e88)
• allow methods in tuning to return headers in sdk_http_response by default (1b1205b)
• Automatically set response type in FunctionDeclaration during the AFC (5ce99df)
• Batches support in Java (5ce13e9)
• enable continuous fine-tuning on a pre-tuned model in the SDK. (e49d350)
• Enable Google Maps tool for Genai. (a4baf3c)
• enable language code for audio transcription config in Live API for Vertex AI (fb034ff)
• Enable R2V and Advanced Controls for Veo on Gemini Developer API (4c42e65)
• enable responseId for Gemini Developer API (4912ff4)
• enable server side MCP and disable all other AFC when server side MCP is configured. (498a2c4)
• Ephemeral token for Gemini Live API support in Java (4ce094b)
• examples (917aee0)
• expose JsonSerializable.stringToJsonNode to help user better use *JsonSchema fields. (35d783b)
• expose session id in Live API (b6d5389)
• genai: add TURN_INCLUDES_AUDIO_ACTIVITY_AND_ALL_VIDEO to TurnCoverage (dbc4c33)
• generate the function_call class's converters (38703c7)
• initial integration (917aee0)
• java local tokenizer (d774185)
• Migrate Go file.create method to use the sdk_http_response field and remove http_headers (7755eab)
• Pass file name to the backend when uploading with a file path (081a9a6)
• rename ComputerUse tool (early access) (4bbba2b)
• return headers for list method in all modules. (45a027b)
• Return response headers for all methods (except streaming methods) (7e8b71b)
• Return response headers for generateContentStream (82a8118)
• set up pom + fix test (917aee0)
• Support API keys for VertexAI mode for Java SDK (826c0dc)
• Support Batches delete (782465d)
• support createEmbeddings in Batches.java (8947f6f)
• support default global location when not using api key with vertexai backend (f9028a7)
• Support different media input in Vertex Live API (7f4c6bf)
• Support distillation tuning (cec1b88)
• support document name in grounding metadata (8273922)
• Support enableWidget feature in GoogleMaps (aefbd5c)
• Support encryption_spec in tuning job creation configuration for GenAI SDK (7a4fb51)
• Support exclude_domains in Google Search and Enterprise Web Search (e975d28)
• support extraBody in HttpOptions class (036bac8)
• support Function call argument streaming for all languages (f310452)
• Support Gemini batch inline request's metadata and add test coverage to safety setting (17033b3)
• Support GenerateVideosSource for Veo GenerateVideos (c26af63)
• Support http headers in GenerateContentResponse (5282774)
• support hyperparameters in distillation tuning (fdedc74)
• Support image recontext on Vertex (e7de8c8)
• Support Imagen 4 Ingredients on Vertex (b5eed8d)
• Support Imagen image segmentation on Vertex (e2a561b)
• Support include_server_side_tool_invocations for genai. (d92fdb1)
• support jailbreak in HarmCategory and BlockedReason (3dab40b)
• Support lists as function parameters in AFC (fixes #527) (452d2e5)
• support multi speaker for Vertex AI (c50e47b)
• Support multimodal embedding for Gemini Embedding 2.0 and support MaaS models in Models.embed_content() (Vertex AI API) (23a7913)
• Support new enum types for UrlRetrievalStatus (cb27222)
• Support OSS Tuning in GenAI SDK (868d8ed)
• Support overriding the max read length in the JSON parser (29d2fca)
• Support rendered_parts in GroundingSupport (6c35249)
• support response headers in Go for all methods. (222b41e)
• Support retries in API requests (3d5de00)
• support tunings.cancel in the genai SDK for Python, Java, JS, and Go (9982251)
• support upload to file search stores (7862ce3)
• Support Veo 2 Editing on Vertex (d401d3c)
• Support Veo 2 Reference Images to Video Generation on Vertex (2f5580f)
• Support video extension for Veo on Gemini Developer API (b398509)
• Update comments in releases.txt (1368773)
• Update data types from discovery doc. (850d527)
• Update data types from discovery doc. (277fd4d)
• Update data types from discovery doc. (fb1f53d)
• Update data types from discovery doc. (6444fb0)
• Update data types from discovery doc. (36ba543)
• Update data types from discovery doc. (42e76ee)
• update release-please configuration to enable version updating across all files (d2f3c0e)
• voice activity support (5ffcf2b)

Bug Fixes

• 'No SLF4J providers were found' on examples (97fa11f)
• Add metadata in batch inlined response (c80dd07)
• Add missing fields to the model types (7b7b41f)
• Defer loading ADC when credentials is provided explicitly in Live API (a540614)
• disable AFC when there are AFC incompatible tool presented. (6099d87)
• Enable id field in FunctionCall for Vertex AI. (3773fe7)
• Ensure Live server message are properly converted (206dc88)
• Expose JOB_STATE_RUNNING and JOB_STATE_EXPIRED for Gemini Batches states (c5b4fdf)
• Fix base_steps parameter for recontext_image (85aaa79)
• fix broken javadoc by switching to dokka. (4b76baf)
• Fix models.list() filter parameter (123ada5)
• Fix service_tier enums. (d4641e4)
• Fix the bug that files.create doesn't return the upload URL correctly (eb40c5f)
• initialization of pre_tuned_model_checkpoint_id from tuning config. (c293633)
• initialize comprehensive and curated history in constructor (f74a426)
• internal workflow change (b80b142)
• live: Enhance security by moving api key from query parameters to header (e48c7f1)
• Make Apache HTTP Components an optional Maven dependency in GenAI. (a01e464)
• Make async generateContent and generateContentStream truly non-blocking (5cb18fd)
• Make async methods in Batches module truly non-blocking (f2ae75a)
• Make async methods in Caches, Tuning, and Operations modules truly non-blocking (db56239)
• Make async methods in Models module truly non-blocking (c205d01)
• Match the versions of the Java Protobuf and Google API Common dependencies (com.google.protobuf:protobuf-java:3.25.5 & com.google.api:api-common:2.45.0) with Google Cloud Java SDKs. (6c37f58)
• only run unit tests in github action (9b2861b)
• Pager throws an exception if list request returns nothing(correct behavior is returning a Pager without any items in it) (0a2301b)
• Remove default timeout (d1f6201)
• Remove duplicate JavaTimeModule in JsonSerializable (a7dbd4c)
• Replace System .err & .out with correct (JUL) Logging in GenAI AsyncLive. (1f756d6)
• support us region routing (3296bfe)
• update Live API audio example with better interruption handling (cad8df9)
• use NonClosingDelegatingHttpClient (917aee0)

Documentation

• Add an example for ClientOptions usage (0a8a26e)
• add blank line before version update marker (82616c2)
• Add docstring for classes and fields that are not supported in Gemini or Vertex API (7a03dac)
• Add docstring for enum classes that are not supported in Gemini or Vertex API (830a12f)
• Add documentation for the retry behavior (4fbcf51)
• Add Imagen and Veo to README (cc0a0aa)
• Add latest models features in README (a2eccaf)
• Add README for Files API (6d206aa)
• explain how to run tests (4f6a811)
• mark Client as thread safe and Chat as not thread safe (be3e50e)
• Recommend using response_json_schema in error messages and docstrings. (6b952e9)
• Refactor model IDs into a Constants class (dacd787)
• Refactor/update docstrings for Imagen and Veo (2470101)
• Remove deprecated product recontext model samples from docstrings (6f73c4a)
• Remove no-op PRs from changelog. (c498c8b)
• Update generated video resolution config docstring (9a2ced8)
• Update Java SDK README spacing (82616c2)
• Update README with latest features in Client (dcf70cc)
• update TokensInfo docstring (48eba7f)
• Update Virtual Try-On model id in samples and docstrings (e349635)

This PR was generated with Release Please. See documentation.