diff --git a/python/packages/main/agent_framework/_types.py b/python/packages/main/agent_framework/_types.py
index d3b3eab044..cf6aa1964f 100644
--- a/python/packages/main/agent_framework/_types.py
+++ b/python/packages/main/agent_framework/_types.py
@@ -1870,10 +1870,17 @@ class ChatOptions(AFBaseModel):
         # No tool choice if no tools are defined
         if self.tools is None or len(self.tools) == 0:
             default_exclude.add("tool_choice")
+        # No metadata and logit bias if they are empty
+        # Prevents 400 error
+        if not self.logit_bias:
+            default_exclude.add("logit_bias")
+        if not self.metadata:
+            default_exclude.add("metadata")
+
         merged_exclude = default_exclude if exclude is None else default_exclude | set(exclude)
 
         settings = self.model_dump(exclude_none=True, by_alias=by_alias, exclude=merged_exclude)
-        settings = {k: v for k, v in settings.items() if v}
+        settings = {k: v for k, v in settings.items() if v is not None}
         settings.update(self.additional_properties)
         for key in merged_exclude:
             settings.pop(key, None)
diff --git a/python/packages/main/tests/main/test_types.py b/python/packages/main/tests/main/test_types.py
index 094d4ea494..87bf31f138 100644
--- a/python/packages/main/tests/main/test_types.py
+++ b/python/packages/main/tests/main/test_types.py
@@ -1138,6 +1138,59 @@ def test_chat_options_tool_choice_dict_mapping(ai_tool):
     assert settings["tool_choice"] == "required"
 
 
+def test_chat_options_to_provider_settings_with_falsy_values():
+    """Test that falsy values (except None) are included in provider settings."""
+    options = ChatOptions(
+        temperature=0.0,  # falsy but not None
+        top_p=0.0,  # falsy but not None
+        presence_penalty=False,  # falsy but not None
+        frequency_penalty=None,  # None - should be excluded
+        additional_properties={"empty_string": "", "zero": 0, "false_flag": False, "none_value": None},
+    )
+
+    settings = options.to_provider_settings()
+
+    # Falsy values that are not None should be included
+    assert "temperature" in settings
+    assert isinstance(settings["temperature"], float)
+    assert settings["temperature"] == 0.0
+    assert "top_p" in settings
+    assert isinstance(settings["top_p"], float)
+    assert settings["top_p"] == 0.0
+    assert "presence_penalty" in settings
+    assert isinstance(settings["presence_penalty"], float)  # converted to float
+    assert settings["presence_penalty"] == 0.0
+
+    # None values should be excluded
+    assert "frequency_penalty" not in settings
+
+    # Additional properties - falsy values should always be included
+    assert "empty_string" in settings
+    assert settings["empty_string"] == ""
+    assert "zero" in settings
+    assert settings["zero"] == 0
+    assert "false_flag" in settings
+    assert settings["false_flag"] is False
+    assert "none_value" in settings
+    assert settings["none_value"] is None
+
+
+def test_chat_options_empty_logit_bias_and_metadata_excluded():
+    """Test that empty logit_bias and metadata are excluded from provider settings."""
+    options = ChatOptions(
+        ai_model_id="gpt-4o",
+        logit_bias={},  # empty dict should be excluded
+        metadata={},  # empty dict should be excluded
+    )
+
+    settings = options.to_provider_settings()
+
+    # Empty logit_bias and metadata should be excluded
+    assert "logit_bias" not in settings
+    assert "metadata" not in settings
+    assert settings["model"] == "gpt-4o"
+
+
 # region AgentRunResponse