diff --git a/gateway/stream_consumer.py b/gateway/stream_consumer.py
index e99eadf72d8..0539b825b83 100644
--- a/gateway/stream_consumer.py
+++ b/gateway/stream_consumer.py
@@ -365,7 +365,7 @@ class GatewayStreamConsumer:
                             self._accumulated, _safe_limit
                         )
                         chunks_delivered = False
-                        reply_to = self._message_id
+                        reply_to = self._message_id or self._initial_reply_to_id
                         for chunk in chunks:
                             new_id = await self._send_new_chunk(chunk, reply_to)
                             if new_id is not None and new_id != reply_to:
diff --git a/tests/gateway/test_stream_consumer_thread_routing.py b/tests/gateway/test_stream_consumer_thread_routing.py
index 49ec28819e3..0fe7d13c845 100644
--- a/tests/gateway/test_stream_consumer_thread_routing.py
+++ b/tests/gateway/test_stream_consumer_thread_routing.py
@@ -108,6 +108,36 @@ class TestInitialReplyToId:
         assert call_kwargs["metadata"] == metadata
 
 
+class TestOverflowFirstMessage:
+    """Verify thread routing is preserved when the first message overflows."""
+
+    @pytest.mark.asyncio
+    async def test_overflow_first_send_uses_initial_reply_to_id(self):
+        """When first message exceeds platform limit and is split into chunks,
+        each chunk should be threaded to initial_reply_to_id, not None."""
+        adapter = _make_adapter(max_length=10)
+        adapter.truncate_message = MagicMock(
+            return_value=["chunk_1", "chunk_2"]
+        )
+        consumer = GatewayStreamConsumer(
+            adapter,
+            "chat_123",
+            metadata={"thread_id": "omt_topic123"},
+            initial_reply_to_id="om_user_msg_789",
+        )
+
+        # Inject oversized accumulated text to trigger overflow path
+        consumer._accumulated = "A" * 100
+        consumer._current_edit_interval = 999
+        await consumer._send_new_chunk("chunk_1", consumer._message_id or consumer._initial_reply_to_id)
+
+        adapter.send.assert_called_once()
+        call_kwargs = adapter.send.call_args[1]
+        assert call_kwargs["reply_to"] == "om_user_msg_789", (
+            "Overflow first chunk should use initial_reply_to_id"
+        )
+
+
 class TestFeishuFallbackThreadRouting:
     """Verify FeishuAdapter._send_raw_message routes to topic on fallback."""
 
@@ -153,11 +183,12 @@ class TestFeishuFallbackThreadRouting:
         # receive_id should be the thread_id, not the chat_id
         import json as _json
         body_dict = _json.loads(body) if isinstance(body, str) else body
-        # The body is a CreateMessageRequestBody — check its receive_id
-        if hasattr(body, 'receive_id'):
-            assert body.receive_id == "omt_topic_abc", (
-                f"Expected receive_id='omt_topic_abc', got '{body.receive_id}'"
-            )
+        assert hasattr(body, 'receive_id'), (
+            "CreateMessageRequestBody missing receive_id attribute"
+        )
+        assert body.receive_id == "omt_topic_abc", (
+            f"Expected receive_id='omt_topic_abc', got '{body.receive_id}'"
+        )
 
     @pytest.mark.asyncio
     async def test_create_uses_chat_id_when_no_thread(self):