diff --git a/TTS/vc/models/freevc.py b/TTS/vc/models/freevc.py
index f9e691256e..262fb7504f 100644
--- a/TTS/vc/models/freevc.py
+++ b/TTS/vc/models/freevc.py
@@ -382,7 +382,8 @@ def load_pretrained_speaker_encoder(self):
         """Load pretrained speaker encoder model as mentioned in the paper."""
         logger.info("Loading pretrained speaker encoder model ...")
         self.enc_spk_ex = SpeakerEncoderEx(
-            "https://github.com/coqui-ai/TTS/releases/download/v0.13.0_models/speaker_encoder.pt"
+            "https://github.com/coqui-ai/TTS/releases/download/v0.13.0_models/speaker_encoder.pt",
+            device=self.device
         )
 
     def init_multispeaker(self, config: Coqpit):