diff --git a/libs/partners/xai/langchain_xai/chat_models.py b/libs/partners/xai/langchain_xai/chat_models.py
index 8e5d5c7b948..00a99c4e6a5 100644
--- a/libs/partners/xai/langchain_xai/chat_models.py
+++ b/libs/partners/xai/langchain_xai/chat_models.py
@@ -575,6 +575,21 @@ class ChatXAI(BaseChatOpenAI):  # type: ignore[override]
                 response.citations
             )
 
+        # Unlike OpenAI, xAI reports reasoning tokens < completion tokens. So we assume
+        # they are not counted in output tokens, and we add them here.
+        if (
+            (not self._use_responses_api({}))
+            and (usage_metadata := rtn.generations[0].message.usage_metadata)  # type: ignore[attr-defined]
+            and (
+                reasoning_tokens := usage_metadata.get("output_token_details", {}).get(
+                    "reasoning"
+                )
+            )
+        ):
+            rtn.generations[0].message.usage_metadata["output_tokens"] += (  # type: ignore[attr-defined]
+                reasoning_tokens
+            )
+
         return rtn
 
     def _convert_chunk_to_generation_chunk(
@@ -609,6 +624,19 @@ class ChatXAI(BaseChatOpenAI):  # type: ignore[override]
         ):
             generation_chunk.message.additional_kwargs["citations"] = citations
 
+        # Unlike OpenAI, xAI reports reasoning tokens < completion tokens. So we assume
+        # they are not counted in output tokens, and we add them here.
+        if (
+            generation_chunk
+            and (not self._use_responses_api({}))
+            and (usage_metadata := generation_chunk.message.usage_metadata)  # type: ignore[attr-defined]
+            and (
+                reasoning_tokens := usage_metadata.get("output_token_details", {}).get(
+                    "reasoning"
+                )
+            )
+        ):
+            generation_chunk.message.usage_metadata["output_tokens"] += reasoning_tokens  # type: ignore[attr-defined]
         return generation_chunk
 
     def with_structured_output(
diff --git a/libs/partners/xai/tests/integration_tests/test_chat_models.py b/libs/partners/xai/tests/integration_tests/test_chat_models.py
index 72d17fc9d19..6a192b5dc99 100644
--- a/libs/partners/xai/tests/integration_tests/test_chat_models.py
+++ b/libs/partners/xai/tests/integration_tests/test_chat_models.py
@@ -37,6 +37,15 @@ def test_reasoning(output_version: Literal["", "v1"]) -> None:
     assert response.content
     assert response.additional_kwargs["reasoning_content"]
 
+    ## Check output tokens
+    usage_metadata = response.usage_metadata
+    assert usage_metadata
+    reasoning_tokens = usage_metadata.get("output_token_details", {}).get("reasoning")
+    total_tokens = usage_metadata.get("output_tokens")
+    assert total_tokens
+    assert reasoning_tokens
+    assert total_tokens > reasoning_tokens
+
     # Test streaming
     full: BaseMessageChunk | None = None
     for chunk in chat_model.stream(input_message):
@@ -44,6 +53,15 @@ def test_reasoning(output_version: Literal["", "v1"]) -> None:
     assert isinstance(full, AIMessageChunk)
     assert full.additional_kwargs["reasoning_content"]
 
+    ## Check output tokens
+    usage_metadata = full.usage_metadata
+    assert usage_metadata
+    reasoning_tokens = usage_metadata.get("output_token_details", {}).get("reasoning")
+    total_tokens = usage_metadata.get("output_tokens")
+    assert total_tokens
+    assert reasoning_tokens
+    assert total_tokens > reasoning_tokens
+
     # Check that we can access reasoning content blocks
     assert response.content_blocks
     reasoning_content = (