diff --git a/src/any_llm/providers/bedrock/bedrock.py b/src/any_llm/providers/bedrock/bedrock.py
index cce7594f..8feeefac 100644
--- a/src/any_llm/providers/bedrock/bedrock.py
+++ b/src/any_llm/providers/bedrock/bedrock.py
@@ -37,7 +37,7 @@ class BedrockProvider(AnyLLM):
     SUPPORTS_COMPLETION = True
     SUPPORTS_RESPONSES = False
     SUPPORTS_COMPLETION_REASONING = True
-    SUPPORTS_COMPLETION_IMAGE = False
+    SUPPORTS_COMPLETION_IMAGE = True
     SUPPORTS_COMPLETION_PDF = False
     SUPPORTS_EMBEDDING = True
     SUPPORTS_LIST_MODELS = True
diff --git a/src/any_llm/providers/bedrock/utils.py b/src/any_llm/providers/bedrock/utils.py
index 27a0650f..4b16f64d 100644
--- a/src/any_llm/providers/bedrock/utils.py
+++ b/src/any_llm/providers/bedrock/utils.py
@@ -114,15 +114,82 @@ def _convert_messages(messages: list[dict[str, Any]]) -> tuple[list[dict[str, An
             if bedrock_message:
                 formatted_messages.append(bedrock_message)
         else:  # user messages
+            if isinstance(message.get("content"), list):
+                # Handle messages with structured content (e.g., images)
+                content = _convert_images(message["content"])
+            else:
+                # Handle simple text messages
+                content = [{"text": message["content"]}]
             formatted_messages.append(
                 {
                     "role": message["role"],
-                    "content": [{"text": message["content"]}],
+                    "content": content,
                 }
             )
 
     return system_message, formatted_messages
 
+def _convert_images(content: list[dict[str, Any]]) -> list[dict[str, Any]]:
+    """Convert images from OpenAI format to AWS Bedrock format.
+
+    - Parse the "content" field block by block
+    - Convert image blocks to Bedrock format
+    - Support both URL and base64 encoded images
+    """
+    converted_content = []
+    for block in content:
+        if block.get("type") == "image_url":
+            converted_block: dict[str, Any] = {"image": {}}
+            url = block.get("image_url", {}).get("url", "")
+
+            if url.startswith("data:image/"):
+                # Handle base64 encoded images
+                mime_part = url[5:]  # Remove "data:"
+                semi_idx = mime_part.find(";")
+                if semi_idx != -1:
+                    media_type = mime_part[:semi_idx]
+                    # Extract format from media type (e.g., "image/png" -> "png")
+                    format_type = media_type.split("/")[1] if "/" in media_type else "png"
+                    base64_data = url.split("base64,", 1)[1] if "base64," in url else ""
+                    converted_block["image"] = {
+                        "format": format_type,
+                        "source": {
+                            "type": "base64",
+                            "data": base64_data,
+                        }
+                    }
+                else:
+                    # Fallback if parsing fails
+                    converted_block["image"] = {
+                        "format": "png",
+                        "source": {
+                            "type": "base64",
+                            "data": url.split("base64,", 1)[1] if "base64," in url else "",
+                        }
+                    }
+            else:
+                # Handle URL images - need to determine format from URL or default to png
+                format_type = "png"  # Default format
+                if url.lower().endswith((".jpg", ".jpeg")):
+                    format_type = "jpeg"
+                elif url.lower().endswith(".png"):
+                    format_type = "png"
+                elif url.lower().endswith(".gif"):
+                    format_type = "gif"
+                elif url.lower().endswith(".webp"):
+                    format_type = "webp"
+
+                converted_block["image"] = {
+                    "format": format_type,
+                    "source": {
+                        "type": "url",
+                        "url": url,
+                    }
+                }
+            converted_content.append(converted_block)
+        else:
+            converted_content.append(block)
+    return converted_content
 
 def _convert_tool_result(message: dict[str, Any]) -> dict[str, Any] | None:
     """Convert OpenAI tool result format to AWS Bedrock format."""
diff --git a/tests/unit/providers/test_aws_provider.py b/tests/unit/providers/test_aws_provider.py
index a14e26b0..8fb05a0a 100644
--- a/tests/unit/providers/test_aws_provider.py
+++ b/tests/unit/providers/test_aws_provider.py
@@ -150,3 +150,36 @@ def test_embedding_list_of_strings() -> None:
         assert response.data[1].index == 1
         assert response.usage.prompt_tokens == 11
         assert response.usage.total_tokens == 11
+
+
+def test_completion_with_images() -> None:
+    """Test that completion correctly processes messages with images."""
+    model_id = "model-id"
+    messages = [
+        {
+            "role": "user",
+            "content": [
+                {"type": "text", "text": "Some question about these images."},
+                {"type": "image_url", "image_url": {"url": "https://example.com/a.png"}},
+                {"type": "image_url", "image_url": {"url": "data:image/jpeg;base64,qwertyuiopasdfghjklzxcvbnm"}},
+            ],
+        }
+    ]
+
+    with mock_aws_provider() as mock_boto3_client:
+        provider = BedrockProvider(api_key="test_key")
+        provider._completion(CompletionParams(model_id=model_id, messages=messages))
+
+        mock_boto3_client.return_value.converse.assert_called_once_with(
+            messages=[
+                {
+                    "role": "user",
+                    "content": [
+                        {"type": "text", "text": "Some question about these images."},
+                        {"image": {"format": "png", "source": {"type": "url", "url": "https://example.com/a.png"}}},
+                        {"image": {"format": "jpeg", "source": {"type": "base64", "data": "qwertyuiopasdfghjklzxcvbnm"}}},
+                    ],
+                }
+            ],
+            modelId=model_id,
+        )