feat: Add GoogleSearch tool type

Google APIs · copybara-github · commit 130b113520c1 · 2025-01-07T16:31:31.000-08:00
feat: Add return type `Schema response` to function declarations
feat: Add id to FunctionCall and FunctionResponse
feat: Add civic_integrity toggle to generation_config
feat: Add response_modalities to generation_config
feat: Add voice_config to generation_config
docs: Update safety filter list to include civic_integrity
feat: Add image_safety block_reason + finish_reason

PiperOrigin-RevId: 713080033
diff --git a/google/ai/generativelanguage/v1beta/BUILD.bazel b/google/ai/generativelanguage/v1beta/BUILD.bazel
@@ -370,6 +370,7 @@ load(
 
 csharp_proto_library(
     name = "generativelanguage_csharp_proto",
+    extra_opts = [],
     deps = [":generativelanguage_proto"],
 )
 
diff --git a/google/ai/generativelanguage/v1beta/content.proto b/google/ai/generativelanguage/v1beta/content.proto
@@ -188,6 +188,10 @@ message CodeExecutionResult {
 // external systems to perform an action, or set of actions, outside of
 // knowledge and scope of the model.
 message Tool {
+  // GoogleSearch tool type.
+  // Tool to support Google Search in Model. Powered by Google.
+  message GoogleSearch {}
+
   // Optional. A list of `FunctionDeclarations` available to the model that can
   // be used for function calling.
   //
@@ -210,6 +214,10 @@ message Tool {
 
   // Optional. Enables the model to execute code as part of generation.
   CodeExecution code_execution = 3 [(google.api.field_behavior) = OPTIONAL];
+
+  // Optional. GoogleSearch tool type.
+  // Tool to support Google Search in Model. Powered by Google.
+  GoogleSearch google_search = 4 [(google.api.field_behavior) = OPTIONAL];
 }
 
 // Tool to retrieve public web data for grounding, powered by Google.
@@ -308,12 +316,21 @@ message FunctionDeclaration {
   // names are case sensitive. Schema Value: the Schema defining the type used
   // for the parameter.
   optional Schema parameters = 3 [(google.api.field_behavior) = OPTIONAL];
+
+  // Optional. Describes the output from this function in JSON Schema format.
+  // Reflects the Open API 3.03 Response Object. The Schema defines the type
+  // used for the response value of the function.
+  optional Schema response = 4 [(google.api.field_behavior) = OPTIONAL];
 }
 
 // A predicted `FunctionCall` returned from the model that contains
 // a string representing the `FunctionDeclaration.name` with the
 // arguments and their values.
 message FunctionCall {
+  // Optional. The unique id of the function call. If populated, the client to
+  // execute the `function_call` and return the response with the matching `id`.
+  string id = 3 [(google.api.field_behavior) = OPTIONAL];
+
   // Required. The name of the function to call.
   // Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum
   // length of 63.
@@ -330,6 +347,10 @@ message FunctionCall {
 // the model. This should contain the result of a`FunctionCall` made
 // based on model prediction.
 message FunctionResponse {
+  // Optional. The id of the function call this response is for. Populated by
+  // the client to match the corresponding function call `id`.
+  string id = 3 [(google.api.field_behavior) = OPTIONAL];
+
   // Required. The name of the function to call.
   // Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum
   // length of 63.
diff --git a/google/ai/generativelanguage/v1beta/file.proto b/google/ai/generativelanguage/v1beta/file.proto
@@ -28,6 +28,7 @@ option java_outer_classname = "FileProto";
 option java_package = "com.google.ai.generativelanguage.v1beta";
 
 // A file uploaded to the API.
+// Next ID: 15
 message File {
   option (google.api.resource) = {
     type: "generativelanguage.googleapis.com/File"
diff --git a/google/ai/generativelanguage/v1beta/generative_service.proto b/google/ai/generativelanguage/v1beta/generative_service.proto
@@ -148,7 +148,7 @@ enum TaskType {
 message GenerateContentRequest {
   // Required. The name of the `Model` to use for generating the completion.
   //
-  // Format: `name=models/{model}`.
+  // Format: `models/{model}`.
   string model = 1 [
     (google.api.field_behavior) = REQUIRED,
     (google.api.resource_reference) = {
@@ -200,8 +200,8 @@ message GenerateContentRequest {
   // `SafetyCategory` provided in the list, the API will use the default safety
   // setting for that category. Harm categories HARM_CATEGORY_HATE_SPEECH,
   // HARM_CATEGORY_SEXUALLY_EXPLICIT, HARM_CATEGORY_DANGEROUS_CONTENT,
-  // HARM_CATEGORY_HARASSMENT are supported. Refer to the
-  // [guide](https://ai.google.dev/gemini-api/docs/safety-settings)
+  // HARM_CATEGORY_HARASSMENT, HARM_CATEGORY_CIVIC_INTEGRITY are supported.
+  // Refer to the [guide](https://ai.google.dev/gemini-api/docs/safety-settings)
   // for detailed information on available safety settings. Also refer to the
   // [Safety guidance](https://ai.google.dev/gemini-api/docs/safety-guidance) to
   // learn how to incorporate safety considerations in your AI applications.
@@ -223,9 +223,45 @@ message GenerateContentRequest {
   ];
 }
 
+// The configuration for the prebuilt speaker to use.
+message PrebuiltVoiceConfig {
+  // The name of the preset voice to use.
+  optional string voice_name = 1;
+}
+
+// The configuration for the voice to use.
+message VoiceConfig {
+  // The configuration for the speaker to use.
+  oneof voice_config {
+    // The configuration for the prebuilt voice to use.
+    PrebuiltVoiceConfig prebuilt_voice_config = 1;
+  }
+}
+
+// The speech generation config.
+message SpeechConfig {
+  // The configuration for the speaker to use.
+  VoiceConfig voice_config = 1;
+}
+
 // Configuration options for model generation and outputs. Not all parameters
 // are configurable for every model.
 message GenerationConfig {
+  // Supported modalities of the response.
+  enum Modality {
+    // Default value.
+    MODALITY_UNSPECIFIED = 0;
+
+    // Indicates the model should return text.
+    TEXT = 1;
+
+    // Indicates the model should return images.
+    IMAGE = 2;
+
+    // Indicates the model should return audio.
+    AUDIO = 3;
+  }
+
   // Optional. Number of generated responses to return.
   //
   // Currently, this value can only be set to 1. If unset, this will default
@@ -343,6 +379,27 @@ message GenerationConfig {
   // This sets the number of top logprobs to return at each decoding step in the
   // [Candidate.logprobs_result][google.ai.generativelanguage.v1beta.Candidate.logprobs_result].
   optional int32 logprobs = 18 [(google.api.field_behavior) = OPTIONAL];
+
+  // Optional. Enables enhanced civic answers. It may not be available for all
+  // models.
+  optional bool enable_enhanced_civic_answers = 19
+      [(google.api.field_behavior) = OPTIONAL];
+
+  // Optional. The requested modalities of the response. Represents the set of
+  // modalities that the model can return, and should be expected in the
+  // response. This is an exact match to the modalities of the response.
+  //
+  // A model may have multiple combinations of supported modalities. If the
+  // requested modalities do not match any of the supported combinations, an
+  // error will be returned.
+  //
+  // An empty list is equivalent to requesting only text.
+  repeated Modality response_modalities = 20
+      [(google.api.field_behavior) = OPTIONAL];
+
+  // Optional. The speech generation config.
+  optional SpeechConfig speech_config = 21
+      [(google.api.field_behavior) = OPTIONAL];
 }
 
 // Configuration for retrieving grounding content from a `Corpus` or
@@ -401,6 +458,9 @@ message GenerateContentResponse {
 
       // Prompt was blocked due to prohibited content.
       PROHIBITED_CONTENT = 4;
+
+      // Candidates blocked due to unsafe image generation content.
+      IMAGE_SAFETY = 5;
     }
 
     // Optional. If set, the prompt was blocked and no candidates are returned.
@@ -481,6 +541,10 @@ message Candidate {
 
     // The function call generated by the model is invalid.
     MALFORMED_FUNCTION_CALL = 10;
+
+    // Token generation stopped because generated images contain safety
+    // violations.
+    IMAGE_SAFETY = 11;
   }
 
   // Output only. Index of the candidate in the list of response candidates.

Original file line number	Diff line number	Diff line change
`@@ -370,6 +370,7 @@ load(`
`370`	`370`
`371`	`371`	`csharp_proto_library(`
`372`	`372`	`name = "generativelanguage_csharp_proto",`
	`373`	`+ extra_opts = [],`
`373`	`374`	`deps = [":generativelanguage_proto"],`
`374`	`375`	`)`
`375`	`376`