diff --git a/admin/dashboard/blueprints/contacts.py b/admin/dashboard/blueprints/contacts.py
index c05e243..59a8517 100644
--- a/admin/dashboard/blueprints/contacts.py
+++ b/admin/dashboard/blueprints/contacts.py
@@ -267,7 +267,7 @@ def api_send_message():
                 return jsonify({'success': False, 'message': '未上传文件'})
             file = request.files['file']
             client_msg_id, create_time, new_msg_id = loop.run_until_complete(
-                server.client.send_voice_message(wxid, file.read(), format="MP3")
+                server.client.send_voice_message(wxid, file.read(), format="mp3")
             )
             return jsonify({
                 'success': True,
diff --git a/wechat_ipad/client/message.py b/wechat_ipad/client/message.py
index afe1624..5b4c013 100644
--- a/wechat_ipad/client/message.py
+++ b/wechat_ipad/client/message.py
@@ -280,7 +280,7 @@ class MessageMixin(WechatAPIClientBase):
         else:
             self.error_handler(json_resp)
 
-    async def send_voice_message(self, wxid: str, voice: Union[str, bytes, os.PathLike], format: str = "AMR") -> \
+    async def send_voice_message(self, wxid: str, voice: Union[str, bytes, os.PathLike], format: str = "amr") -> \
             tuple[int, int, int]:
         """发送语音消息。
 
@@ -300,12 +300,12 @@ class MessageMixin(WechatAPIClientBase):
         """
         return await self._queue_message(self._send_voice_message, wxid, voice, format)
 
-    async def _send_voice_message(self, wxid: str, voice: Union[str, bytes, os.PathLike], format: str = "AMR") -> \
+    async def _send_voice_message(self, wxid: str, voice: Union[str, bytes, os.PathLike], format: str = "mar") -> \
             tuple[int, int, int]:
         if not self.wxid:
             raise UserLoggedOut("请先登录")
 
-        elif format not in ["AMR", "WAVE", "MP3", "SILK", "SPEEX"]:
+        elif format not in ["amr", "wave", "mp3", "silk", "speex"]:
             raise ValueError("format must be one of amr, wav, mp3")
 
         # read voice to byte