diff --git a/ml-agents/mlagents/trainers/torch/action_model.py b/ml-agents/mlagents/trainers/torch/action_model.py
index 0b0d53706e..c5de586e4d 100644
--- a/ml-agents/mlagents/trainers/torch/action_model.py
+++ b/ml-agents/mlagents/trainers/torch/action_model.py
@@ -164,6 +164,9 @@ def get_action_out(self, inputs: torch.Tensor, masks: torch.Tensor) -> torch.Ten
         if self.action_spec.continuous_size > 0 and dists.continuous is not None:
             continuous_out = dists.continuous.exported_model_output()
             action_out_deprecated = dists.continuous.exported_model_output()
+            if self._clip_action_on_export:
+                continuous_out = torch.clamp(continuous_out, -3, 3) / 3
+                action_out_deprecated = torch.clamp(action_out_deprecated, -3, 3) / 3
         if self.action_spec.discrete_size > 0 and dists.discrete is not None:
             discrete_out_list = [
                 discrete_dist.exported_model_output()