diff --git a/contrib/test_bench.py b/contrib/test_bench.py
index 3fa77d46..b1baf024 100644
--- a/contrib/test_bench.py
+++ b/contrib/test_bench.py
@@ -4,10 +4,8 @@
 
 from yente.data.entity import Entity
 from yente.data.common import EntityExample
-from yente.search.base import close_es
 from yente.search.queries import entity_query
 from yente.search.search import search_entities, result_entities
-from yente.data.entity import Entity
 from yente.scoring import score_results
 from yente.routers.util import get_dataset
 
@@ -34,7 +32,7 @@ async def test_example():
         print(ent.id, ent.caption, ent.schema.name)
 
     algorithm = get_algorithm("name-based")
-    scored = score_results(
+    total, scored = score_results(
         algorithm,
         entity,
         ents,
@@ -43,11 +41,9 @@ async def test_example():
         limit=LIMIT,
     )
 
-    print("\n\nSCORED RESULTS:")
+    print("\n\nSCORED RESULTS [%d]:" % total)
     for res in scored:
         print(res.id, res.caption, res.schema_, res.score)
 
-    await close_es()
-
 
 asyncio.run(test_example())
diff --git a/yente/routers/match.py b/yente/routers/match.py
index b6cb1100..b9653817 100644
--- a/yente/routers/match.py
+++ b/yente/routers/match.py
@@ -6,10 +6,10 @@
 from yente.logs import get_logger
 from yente.data.common import ErrorResponse
 from yente.data.common import EntityMatchQuery, EntityMatchResponse, EntityExample
-from yente.data.common import EntityMatches
+from yente.data.common import EntityMatches, TotalSpec
 from yente.provider import SearchProvider, get_provider
 from yente.search.queries import entity_query, FilterDict
-from yente.search.search import search_entities, result_entities, result_total
+from yente.search.search import search_entities, result_entities
 from yente.data.entity import Entity
 from yente.util import limit_window
 from yente.scoring import score_results
@@ -173,7 +173,7 @@ async def match(
 
     for (name, entity), resp in zip(entities, results):
         ents = result_entities(resp)
-        scored = score_results(
+        total, scored = score_results(
             algorithm_type,
             entity,
             ents,
@@ -182,17 +182,16 @@ async def match(
             limit=limit,
             weights=match.weights,
         )
-        total = result_total(resp)
         log.info(
             f"/match/{ds.name}",
             action="match",
             schema=entity.schema.name,
-            results=total.value,
+            results=total,
         )
         responses[name] = EntityMatches(
             status=200,
             results=scored,
-            total=total,
+            total=TotalSpec(value=total, relation="eq"),
             query=EntityExample.model_validate(entity.to_dict()),
         )
     response.headers["x-batch-size"] = str(len(responses))
diff --git a/yente/routers/reconcile.py b/yente/routers/reconcile.py
index d9e30540..20b58941 100644
--- a/yente/routers/reconcile.py
+++ b/yente/routers/reconcile.py
@@ -188,13 +188,13 @@ async def reconcile_query(
     resp = await search_entities(provider, query, limit=limit, offset=offset)
     algorithm_ = get_algorithm_by_name(algorithm)
     entities = result_entities(resp)
-    scoreds = [s for s in score_results(algorithm_, proxy, entities, limit=limit)]
+    total, scoreds = score_results(algorithm_, proxy, entities, limit=limit)
     results = [FreebaseScoredEntity.from_scored(s) for s in scoreds]
     log.info(
         f"/reconcile/{dataset.name}",
         action="reconcile",
         schema=proxy.schema.name,
-        results=result_total(resp).value,
+        matches=total,
     )
     return name, FreebaseEntityResult(result=results)
 
diff --git a/yente/scoring.py b/yente/scoring.py
index 8fcc2287..a8ef96d6 100644
--- a/yente/scoring.py
+++ b/yente/scoring.py
@@ -1,4 +1,4 @@
-from typing import Iterable, List, Optional, Type, Dict
+from typing import Iterable, List, Optional, Type, Dict, Tuple
 from nomenklatura.matching.types import ScoringAlgorithm
 
 from yente import settings
@@ -14,7 +14,7 @@ def score_results(
     cutoff: float = 0.0,
     limit: Optional[int] = None,
     weights: Dict[str, float] = {},
-) -> List[ScoredEntityResponse]:
+) -> Tuple[int, List[ScoredEntityResponse]]:
     scored: List[ScoredEntityResponse] = []
     matches = 0
     for proxy in results:
@@ -29,4 +29,4 @@ def score_results(
     scored = sorted(scored, key=lambda r: r.score, reverse=True)
     if limit is not None:
         scored = scored[:limit]
-    return scored
+    return matches, scored