diff --git a/integrations/anthropic/src/haystack_integrations/components/generators/anthropic/chat/chat_generator.py b/integrations/anthropic/src/haystack_integrations/components/generators/anthropic/chat/chat_generator.py
index 43b50495c..56a740146 100644
--- a/integrations/anthropic/src/haystack_integrations/components/generators/anthropic/chat/chat_generator.py
+++ b/integrations/anthropic/src/haystack_integrations/components/generators/anthropic/chat/chat_generator.py
@@ -1,4 +1,3 @@
-import dataclasses
 import json
 from typing import Any, Callable, ClassVar, Dict, List, Optional, Union
 
@@ -275,8 +274,16 @@ def _convert_to_anthropic_format(self, messages: List[ChatMessage]) -> List[Dict
         """
         anthropic_formatted_messages = []
         for m in messages:
-            message_dict = dataclasses.asdict(m)
-            formatted_message = {k: v for k, v in message_dict.items() if k in {"role", "content"} and v}
+            message_dict = m.to_dict()
+            formatted_message = {}
+
+            # legacy format
+            if "role" in message_dict and "content" in message_dict:
+                formatted_message = {k: v for k, v in message_dict.items() if k in {"role", "content"} and v}
+            # new format
+            elif "_role" in message_dict and "_content" in message_dict:
+                formatted_message = {"role": m.role.value, "content": m.text}
+
             if m.is_from(ChatRole.SYSTEM):
                 # system messages are treated differently and MUST be in the format expected by the Anthropic API
                 # remove role and content from the message dict, add type and text
diff --git a/integrations/anthropic/tests/test_chat_generator.py b/integrations/anthropic/tests/test_chat_generator.py
index 9a111fc9d..69b3265aa 100644
--- a/integrations/anthropic/tests/test_chat_generator.py
+++ b/integrations/anthropic/tests/test_chat_generator.py
@@ -421,6 +421,8 @@ def test_prompt_caching(self, cache_enabled):
         assert len(result["replies"]) == 1
         token_usage = result["replies"][0].meta.get("usage")
 
+        print(token_usage)
+
         if cache_enabled:
             # either we created cache or we read it (depends on how you execute this integration test)
             assert (
@@ -428,5 +430,5 @@ def test_prompt_caching(self, cache_enabled):
                 or token_usage.get("cache_read_input_tokens") > 1024
             )
         else:
-            assert "cache_creation_input_tokens" not in token_usage
-            assert "cache_read_input_tokens" not in token_usage
+            assert token_usage["cache_creation_input_tokens"] == 0
+            assert token_usage["cache_read_input_tokens"] == 0