test(cost): assert jina rerank spend at the registry rate

litellm/llms/jina_ai/rerank/transformation.py

@@ -157,9 +157,6 @@ class JinaAIRerankConfig(BaseRerankConfig):
         billed_units: RerankBilledUnits | None = None,
         model_info: ModelInfo | None = None,
     ) -> tuple[float, float]:
-        """
-        Jina AI reranker is priced at $0.000000018 per token.
-        """
         if (
             model_info is None
             or "input_cost_per_token" not in model_info

tests/test_litellm/test_cost_calculator.py

@@ -18,6 +18,7 @@ from litellm.cost_calculator import (
     response_cost_calculator,
 )
 from litellm.types.llms.openai import OpenAIRealtimeStreamList
+from litellm.types.rerank import RerankResponse
 from litellm.types.utils import (
     CacheCreationTokenDetails,
     ModelInfo,
@@ -128,6 +129,22 @@ def test_completion_cost_uses_response_model_for_dynamic_routing(_local_model_co
     assert cost > 0, "Cost should be calculated using response model"
 
 
+def test_jina_rerank_bills_total_tokens_at_input_rate_only(_local_model_cost_map):
+    response: Final = RerankResponse(
+        id="rerank-1",
+        results=[{"index": 0, "relevance_score": 0.9}],
+        meta={"billed_units": {"total_tokens": 1000}},
+    )
+
+    cost: Final = completion_cost(
+        completion_response=response,
+        model="jina_ai/jina-reranker-v2-base-multilingual",
+        call_type="rerank",
+    )
+
+    assert cost == pytest.approx(1000 * 5e-08)
+
+
 def test_cost_calculator_with_response_cost_in_additional_headers():
     class MockResponse(BaseModel):
         _hidden_params = {