diff --git a/sdk/rt/speechmatics/rt/_async_client.py b/sdk/rt/speechmatics/rt/_async_client.py
index 892e2549..0aa5864d 100644
--- a/sdk/rt/speechmatics/rt/_async_client.py
+++ b/sdk/rt/speechmatics/rt/_async_client.py
@@ -174,8 +174,7 @@ async def force_end_of_utterance(self, timestamp: Optional[float] = None) -> Non
         transcript to be sent to the client early.
 
         Takes an optional timestamp parameter to specify a marker for the engine
-        to use for timing of the end of the utterance. If not provided, the timestamp
-        will be calculated based on the cumulative audio sent to the server.
+        to use for timing of the end of the utterance.
 
         Args:
             timestamp: Optional timestamp for the request.
@@ -184,7 +183,6 @@ async def force_end_of_utterance(self, timestamp: Optional[float] = None) -> Non
             ConnectionError: If the WebSocket connection fails.
             TranscriptionError: If the server reports an error during teardown.
             TimeoutError: If the connection or teardown times out.
-            ValueError: If the audio format does not have an encoding set.
 
         Examples:
             Basic streaming:
@@ -193,9 +191,6 @@ async def force_end_of_utterance(self, timestamp: Optional[float] = None) -> Non
                 ...     await client.send_audio(frame)
                 ...     await client.force_end_of_utterance()
         """
-        if timestamp is None:
-            timestamp = self.audio_seconds_sent
-
         await self.send_message({"message": ClientMessageType.FORCE_END_OF_UTTERANCE, "timestamp": timestamp})
 
     @property
diff --git a/sdk/voice/speechmatics/voice/_client.py b/sdk/voice/speechmatics/voice/_client.py
index bc36f39f..2b0bd3fd 100644
--- a/sdk/voice/speechmatics/voice/_client.py
+++ b/sdk/voice/speechmatics/voice/_client.py
@@ -1687,7 +1687,7 @@ async def _await_forced_eou(self, timeout: float = 1.0) -> None:
             self._forced_eou_active = True
 
             # Send the force EOU and wait for the response
-            await self.force_end_of_utterance()
+            await self.force_end_of_utterance(self.audio_seconds_sent)
             await asyncio.wait_for(eou_received.wait(), timeout=timeout)
 
             # Record the latency