diff --git a/src/rai_core/rai/agents/langchain/core/react_agent.py b/src/rai_core/rai/agents/langchain/core/react_agent.py
index d49fd4617..0439dc211 100644
--- a/src/rai_core/rai/agents/langchain/core/react_agent.py
+++ b/src/rai_core/rai/agents/langchain/core/react_agent.py
@@ -29,7 +29,7 @@
 
 from rai.agents.langchain.core.tool_runner import ToolRunner
 from rai.initialization import get_llm_model
-from rai.messages import SystemMultimodalMessage
+from rai.messages import AIMultimodalMessage, SystemMultimodalMessage
 
 
 class ReActAgentState(TypedDict):
@@ -76,6 +76,7 @@ def llm_node(
         if not isinstance(state["messages"][0], SystemMessage):
             state["messages"].insert(0, SystemMessage(content=system_prompt))
     ai_msg = llm.invoke(state["messages"])
+    ai_msg = AIMultimodalMessage(content=ai_msg.content)
     state["messages"].append(ai_msg)
 
 
diff --git a/src/rai_core/rai/messages/multimodal.py b/src/rai_core/rai/messages/multimodal.py
index b1646f37d..98bb57f57 100644
--- a/src/rai_core/rai/messages/multimodal.py
+++ b/src/rai_core/rai/messages/multimodal.py
@@ -13,10 +13,12 @@
 # limitations under the License.
 
 
+from datetime import datetime
 from typing import Any, Dict, List, Literal, Optional, Union
 
 from langchain_core.messages import AIMessage, HumanMessage, SystemMessage, ToolMessage
 from langchain_core.messages.base import BaseMessage, get_msg_title_repr
+from pydantic import Field
 
 
 class MultimodalMessage(BaseMessage):
@@ -31,6 +33,8 @@ class MultimodalMessage(BaseMessage):
         List of base64 encoded audios.
     """
 
+    timestamp: datetime = Field(default_factory=datetime.now)
+
     images: Optional[List[str]] = None
     audios: Optional[Any] = None
 
@@ -46,7 +50,15 @@ def __init__(
         _content: List[Union[str, Dict[str, Union[Dict[str, str], str]]]] = []
 
         if isinstance(self.content, str):
-            _content.append({"type": "text", "text": self.content})
+            _content.append(
+                {
+                    "type": "text",
+                    "text": "Current time: "
+                    + self.timestamp.isoformat(timespec="seconds")
+                    + "\n"
+                    + self.content,
+                }
+            )
         else:
             raise ValueError("Content must be a string")  # for now, to guarantee compat
 
@@ -63,10 +75,6 @@ def __init__(
             _content.extend(_image_content)
         self.content = _content
 
-    @property
-    def text(self) -> str:
-        return self.content[0]["text"]
-
 
 class HumanMultimodalMessage(HumanMessage, MultimodalMessage):
     def __repr_args__(self) -> Any: