fix(caching): forward metadata to embedding in valkey semantic sync path

litellm/caching/valkey_semantic_cache.py

@@ -240,7 +240,7 @@ class ValkeySemanticCache(RedisSemanticCache):
                 print_verbose("No prompt provided for semantic caching")
                 return
 
-            embedding = self._get_embedding(prompt)
+            embedding = self._get_embedding(prompt, metadata=kwargs.get("metadata"))
             self._ensure_index_sync(len(embedding))
 
             doc_key = self._doc_key(key)
@@ -259,7 +259,7 @@ class ValkeySemanticCache(RedisSemanticCache):
                 kwargs.setdefault("metadata", {})["semantic-similarity"] = 0.0
                 return None
 
-            embedding = self._get_embedding(prompt)
+            embedding = self._get_embedding(prompt, metadata=kwargs.get("metadata"))
             self._ensure_index_sync(len(embedding))
 
             search_result = self.sync_client.ft(self.index_name).search(

tests/test_litellm/caching/test_valkey_semantic_cache.py

@@ -353,6 +353,53 @@ async def test_async_get_cache_passes_only_metadata_to_get_async_embedding():
     assert captured["metadata"] == {"user_api_key": "sk-test"}
 
 
+def test_set_cache_passes_metadata_to_get_embedding():
+    """Regression for #32324: the sync path must forward request metadata to the
+    embedding call so embedding spend is attributed, matching the async path and
+    RedisSemanticCache."""
+    sync_client = MagicMock()
+    cache = _make_cache(sync_client=sync_client)
+    captured: dict[str, object] = {}
+
+    def spy_embedding(prompt: str, metadata: dict | None = None) -> list[float]:
+        captured["metadata"] = metadata
+        return [0.1, 0.2, 0.3]
+
+    cache._get_embedding = spy_embedding
+
+    cache.set_cache(
+        key="cache-key",
+        value={"content": "Paris"},
+        messages=[{"role": "user", "content": "What is the capital of France?"}],
+        metadata={"user_api_key": "sk-test"},
+    )
+
+    assert captured["metadata"] == {"user_api_key": "sk-test"}
+
+
+def test_get_cache_passes_metadata_to_get_embedding():
+    """Regression for #32324: the sync get path must also forward metadata."""
+    sync_client = MagicMock()
+    sync_client.ft.return_value.search.return_value = _search_result(0.1)
+    cache = _make_cache(sync_client=sync_client, similarity_threshold=0.8)
+    captured: dict[str, object] = {}
+
+    def spy_embedding(prompt: str, metadata: dict | None = None) -> list[float]:
+        captured["metadata"] = dict(metadata) if metadata is not None else None
+        return [0.1, 0.2, 0.3]
+
+    cache._get_embedding = spy_embedding
+
+    result = cache.get_cache(
+        key="cache-key",
+        messages=[{"role": "user", "content": "capital of France?"}],
+        metadata={"user_api_key": "sk-test"},
+    )
+
+    assert result == {"content": "Paris"}
+    assert captured["metadata"] == {"user_api_key": "sk-test"}
+
+
 @pytest.mark.asyncio
 async def test_async_get_cache_misses_below_threshold():
     async_client = AsyncMock()