feat: Expose llm parser to public v1 proto to prepare for GA

Google APIs · copybara-github · commit cccef4117d3a · 2025-04-30T23:18:54.000-07:00
PiperOrigin-RevId: 753454988
diff --git a/google/cloud/aiplatform/v1/vertex_rag_data.proto b/google/cloud/aiplatform/v1/vertex_rag_data.proto
@@ -347,10 +347,32 @@ message RagFileParsingConfig {
     int32 max_parsing_requests_per_min = 2;
   }
 
+  // Specifies the advanced parsing for RagFiles.
+  message LlmParser {
+    // The name of a LLM model used for parsing.
+    // Format:
+    // * `projects/{project_id}/locations/{location}/publishers/{publisher}/models/{model}`
+    string model_name = 1;
+
+    // The maximum number of requests the job is allowed to make to the
+    // LLM model per minute. Consult
+    // https://cloud.google.com/vertex-ai/generative-ai/docs/quotas
+    // and your document size to set an appropriate value here. If unspecified,
+    // a default value of 5000 QPM would be used.
+    int32 max_parsing_requests_per_min = 2;
+
+    // The prompt to use for parsing. If not specified, a default prompt will
+    // be used.
+    string custom_parsing_prompt = 3;
+  }
+
   // The parser to use for RagFiles.
   oneof parser {
     // The Layout Parser to use for RagFiles.
     LayoutParser layout_parser = 4;
+
+    // The LLM Parser to use for RagFiles.
+    LlmParser llm_parser = 5;
   }
 }
 
