diff --git a/datajunction-server/datajunction_server/api/graphql/dataloaders.py b/datajunction-server/datajunction_server/api/graphql/dataloaders.py
index 60ec9a566..0d7a2df75 100644
--- a/datajunction-server/datajunction_server/api/graphql/dataloaders.py
+++ b/datajunction-server/datajunction_server/api/graphql/dataloaders.py
@@ -370,7 +370,7 @@ def create_extracted_measures_loader(
     returning N metrics, Strawberry batches all per-node loader.load(nr_id)
     calls within a single event-loop tick into one
     batch_load_extracted_measures(ids) — collapsing N sessions + N extractions
-    into 1 session + N extractions sharing parsed_query_cache.
+    into 1 session + N extractions sharing nodes_cache and parent_map.
     """
     return DataLoader(
         load_fn=lambda keys: batch_load_extracted_measures(keys, request),
diff --git a/datajunction-server/datajunction_server/api/graphql/scalars/node.py b/datajunction-server/datajunction_server/api/graphql/scalars/node.py
index acd2d332b..59b53ec6c 100644
--- a/datajunction-server/datajunction_server/api/graphql/scalars/node.py
+++ b/datajunction-server/datajunction_server/api/graphql/scalars/node.py
@@ -393,8 +393,8 @@ async def extracted_measures(
 
         Uses the request-scoped extracted_measures_loader so that a GraphQL
         query returning N metrics batches all extractions into a single
-        session + shared parsed_query_cache (instead of opening N independent
-        sessions via resolver_session).
+        session + shared nodes_cache/parent_map (instead of opening N
+        independent sessions via resolver_session).
         """
         if root.type != NodeType.METRIC:
             return None
diff --git a/datajunction-server/datajunction_server/internal/nodes.py b/datajunction-server/datajunction_server/internal/nodes.py
index abed7d771..0107a3387 100644
--- a/datajunction-server/datajunction_server/internal/nodes.py
+++ b/datajunction-server/datajunction_server/internal/nodes.py
@@ -750,7 +750,7 @@ async def derive_frozen_measures_bulk(
 
       * one query to load all target revisions + their parent chain eagerly,
       * zero-DB `MetricComponentExtractor.extract` calls (via the extractor's
-        `nodes_cache` / `parent_map` / `parsed_query_cache` params), and
+        `nodes_cache` / `parent_map` params), and
       * one batch `SELECT` against FrozenMeasure.name IN (...).
 
     Caller owns the transaction and commit; used by the deployment
@@ -807,8 +807,6 @@ async def derive_frozen_measures_bulk(
                 for grandparent in parent.current.parents:
                     nodes_cache.setdefault(grandparent.name, grandparent)
 
-    parsed_query_cache: dict[str, ast.Query] = {}
-
     # 3. Per-metric extract with caches — zero DB calls in this loop.
     extraction_results: list[tuple[NodeRevision, list]] = []
     for rev in revisions:
@@ -818,7 +816,6 @@ async def derive_frozen_measures_bulk(
             nodes_cache=nodes_cache,
             parent_map=parent_map,
             metric_node=rev.node,
-            parsed_query_cache=parsed_query_cache,
         )
         rev.derived_expression = str(derived_sql)
         extraction_results.append((rev, measures))
diff --git a/datajunction-server/datajunction_server/sql/decompose.py b/datajunction-server/datajunction_server/sql/decompose.py
index c83e10060..57012c5a3 100644
--- a/datajunction-server/datajunction_server/sql/decompose.py
+++ b/datajunction-server/datajunction_server/sql/decompose.py
@@ -2,7 +2,6 @@
 
 import hashlib
 from abc import ABC, abstractmethod
-from copy import deepcopy
 from dataclasses import dataclass
 from typing import cast
 
@@ -696,7 +695,6 @@ async def extract(
         nodes_cache: dict[str, "Node"] | None = None,
         parent_map: dict[str, list[str]] | None = None,
         metric_node: "Node | None" = None,
-        parsed_query_cache: dict[str, ast.Query] | None = None,
         _visited: set[str] | None = None,
     ) -> tuple[list[MetricComponent], ast.Query]:
         """
@@ -714,8 +712,6 @@ async def extract(
                 Required if nodes_cache is provided.
             metric_node: Optional metric Node object.
                 Required if nodes_cache is provided.
-            parsed_query_cache: Optional dict of query_string -> parsed AST.
-                Used to avoid re-parsing the same query multiple times.
         """
         # Use cache if available, otherwise query DB
         if (
@@ -731,18 +727,7 @@ async def extract(
         else:
             metric_data = await self._load_metric_data(session)
 
-        # Helper to parse with cache
-        def cached_parse(query: str) -> ast.Query:
-            if parsed_query_cache is not None:
-                if query not in parsed_query_cache:  # pragma: no cover
-                    parsed_query_cache[query] = parse(query)
-
-                # Return a deep copy to avoid AST mutation issues
-                return deepcopy(parsed_query_cache[query])  # pragma: no cover
-            return parse(query)
-
-        # Parse queries (pure computation, no DB)
-        query_ast = cached_parse(metric_data.query)
+        query_ast = parse(metric_data.query)
 
         # Initialize visited set for cycle detection
         if _visited is None:
@@ -820,12 +805,11 @@ def cached_parse(query: str) -> ast.Query:
                     nodes_cache=nodes_cache,
                     parent_map=parent_map,
                     metric_node=parent_node,
-                    parsed_query_cache=parsed_query_cache,
                     _visited=_visited,
                 )
             else:
                 # True base metric - decompose aggregations
-                base_ast = cached_parse(base_metric.query)
+                base_ast = parse(base_metric.query)
                 base_components, derived_ast = self._extract_base(base_ast)
 
             for comp in base_components: