Amazon Bedrock Runtime Update: You can now reference images and documents stored in Amazon S3 when using InvokeModel and Converse APIs with Amazon Nova Lite and Nova Pro. This enables direct integration of S3-stored multimedia assets in your model requests without manual downloading or base64 encoding.

Author: AWS

.changes/next-release/feature-AmazonBedrockRuntime-bdfbc89.json

@@ -0,0 +1,6 @@
+{
+    "type": "feature",
+    "category": "Amazon Bedrock Runtime",
+    "contributor": "",
+    "description": "You can now reference images and documents stored in Amazon S3 when using InvokeModel and Converse APIs with Amazon Nova Lite and Nova Pro. This enables direct integration of S3-stored multimedia assets in your model requests without manual downloading or base64 encoding."
+}

services/bedrockruntime/src/main/resources/codegen-resources/service-2.json

@@ -987,6 +987,10 @@
         "bytes":{
           "shape":"DocumentSourceBytesBlob",
           "documentation":"<p>The raw bytes for the document. If you use an Amazon Web Services SDK, you don't need to encode the bytes in base64.</p>"
+        },
+        "s3Location":{
+          "shape":"S3Location",
+          "documentation":"<p>The location of a document object in an Amazon S3 bucket. To see which models support S3 uploads, see <a href=\"https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference-supported-models-features.html\">Supported models and features for Converse</a>.</p>"
         }
       },
       "documentation":"<p>Contains the content of a document.</p>",
@@ -1987,6 +1991,10 @@
         "bytes":{
           "shape":"ImageSourceBytesBlob",
           "documentation":"<p>The raw image bytes for the image. If you use an AWS SDK, you don't need to encode the image bytes in base64.</p>"
+        },
+        "s3Location":{
+          "shape":"S3Location",
+          "documentation":"<p>The location of an image object in an Amazon S3 bucket. To see which models support S3 uploads, see <a href=\"https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference-supported-models-features.html\">Supported models and features for Converse</a>.</p>"
         }
       },
       "documentation":"<p>The source for an image.</p>",
@@ -2726,7 +2734,7 @@
           "documentation":"<p>If the bucket belongs to another AWS account, specify that account's ID.</p>"
         }
       },
-      "documentation":"<p>A storage location in an S3 bucket.</p>"
+      "documentation":"<p>A storage location in an Amazon S3 bucket.</p>"
     },
     "S3Uri":{
       "type":"string",
@@ -3248,7 +3256,7 @@
         },
         "s3Location":{
           "shape":"S3Location",
-          "documentation":"<p>The location of a video object in an S3 bucket.</p>"
+          "documentation":"<p>The location of a video object in an Amazon S3 bucket. To see which models support S3 uploads, see <a href=\"https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference-supported-models-features.html\">Supported models and features for Converse</a>.</p>"
         }
       },
       "documentation":"<p>A video source. You can upload a smaller video as a base64-encoded string as long as the encoded file is less than 25MB. You can also transfer videos up to 1GB in size from an S3 bucket.</p>",