fix(cost): bill OCR annotation pages via annotation_cost_per_page

litellm/cost_calculator.py

@@ -1910,12 +1910,15 @@ def ocr_cost(
     if credits is not None and cost_per_credit is not None:
         return cost_per_credit * credits, 0.0
 
-    ocr_cost_per_page: float | None = None
-    if model_info is not None:
-        ocr_cost_per_page = model_info.get("ocr_cost_per_page")
+    ocr_cost_per_page: Final = model_info.get("ocr_cost_per_page") if model_info is not None else None
+    annotation_cost_per_page: Final = model_info.get("annotation_cost_per_page") if model_info is not None else None
+    annotation_rate: Final = annotation_cost_per_page if annotation_cost_per_page is not None else ocr_cost_per_page
 
     pages_processed: Final = response.usage_info.pages_processed
-    if pages_processed is None:
+    annotation_pages: Final = response.usage_info.pages_processed_annotation or 0
+    has_billable_annotation_pages: Final = annotation_rate is not None and annotation_pages > 0
+
+    if pages_processed is None and not has_billable_annotation_pages:
         if cost_per_credit is not None or ocr_cost_per_page is None:
             # Surface missing usage data instead of silently under-reporting
             # cost. The previous behavior raised ValueError; we now return 0.0
@@ -1931,7 +1934,7 @@ def ocr_cost(
             return 0.0, 0.0
         raise ValueError("OCR response pages_processed is None")
 
-    if ocr_cost_per_page is None:
+    if ocr_cost_per_page is None and not has_billable_annotation_pages:
         # No per-page pricing configured. Either the model is on credit-based
         # pricing (and credits weren't returned, so the credit branch above did
         # not match) or the model has no OCR pricing entry at all. Surface a
@@ -1947,8 +1950,9 @@ def ocr_cost(
         )
         return 0.0, 0.0
 
-    total_ocr_processing_cost: Final[float] = ocr_cost_per_page * pages_processed
-    return total_ocr_processing_cost, 0.0
+    ocr_pages_cost: Final = (ocr_cost_per_page or 0.0) * (pages_processed or 0)
+    annotation_pages_cost: Final = (annotation_rate or 0.0) * annotation_pages
+    return ocr_pages_cost + annotation_pages_cost, 0.0
 
 
 def vector_store_search_cost(

litellm/llms/base_llm/ocr/transformation.py

@@ -75,6 +75,7 @@ class OCRUsageInfo(LiteLLMPydanticObjectBase):
     """Usage information from OCR response."""
 
     pages_processed: int | None = None
+    pages_processed_annotation: int | None = None
     credits: float | None = None
     doc_size_bytes: int | None = None
 

tests/test_litellm/llms/mistral/ocr/test_mistral_ocr_cost.py

@@ -24,6 +24,9 @@ OCR3_MODEL = "mistral/mistral-ocr-2512"
 OCR3_COST_PER_PAGE = 0.002
 OCR3_ANNOTATION_COST_PER_PAGE = 0.003
 
+AZURE_DOC_AI_MODEL = "azure_ai/mistral-document-ai-2512"
+AZURE_DOC_AI_COST_PER_PAGE = 0.003
+
 
 def _ocr_response(model: str, pages_processed: int) -> OCRResponse:
     return OCRResponse(
@@ -33,6 +36,14 @@ def _ocr_response(model: str, pages_processed: int) -> OCRResponse:
     )
 
 
+def _annotated_ocr_response(model: str, pages_processed: int | None, annotation_pages: int) -> OCRResponse:
+    return OCRResponse(
+        pages=[],
+        model=model,
+        usage_info=OCRUsageInfo(pages_processed=pages_processed, pages_processed_annotation=annotation_pages),
+    )
+
+
 @pytest.mark.parametrize("model", ["mistral-ocr-4-0", "mistral-ocr-latest"])
 def test_model_info_ocr4_price(model: str) -> None:
     info = litellm.get_model_info(model=f"mistral/{model}", custom_llm_provider="mistral")
@@ -79,3 +90,46 @@ def test_ocr3_cost_scales_with_pages(local_model_cost_map, pages_processed: int)
         call_type="ocr",
     )
     assert cost == pytest.approx(OCR3_COST_PER_PAGE * pages_processed)
+
+
+def test_ocr3_bills_ocr_and_annotation_pages_at_their_own_rates(local_model_cost_map) -> None:
+    cost = completion_cost(
+        completion_response=_annotated_ocr_response("mistral-ocr-2512", 2, 3),
+        model=OCR3_MODEL,
+        custom_llm_provider="mistral",
+        call_type="ocr",
+    )
+    assert cost == pytest.approx(2 * OCR3_COST_PER_PAGE + 3 * OCR3_ANNOTATION_COST_PER_PAGE)
+
+
+def test_ocr3_bills_annotation_only_response(local_model_cost_map) -> None:
+    cost = completion_cost(
+        completion_response=_annotated_ocr_response("mistral-ocr-2512", 0, 3),
+        model=OCR3_MODEL,
+        custom_llm_provider="mistral",
+        call_type="ocr",
+    )
+    assert cost == pytest.approx(3 * OCR3_ANNOTATION_COST_PER_PAGE)
+
+
+def test_ocr3_bills_annotation_pages_when_pages_processed_missing(local_model_cost_map) -> None:
+    cost = completion_cost(
+        completion_response=_annotated_ocr_response("mistral-ocr-2512", None, 4),
+        model=OCR3_MODEL,
+        custom_llm_provider="mistral",
+        call_type="ocr",
+    )
+    assert cost == pytest.approx(4 * OCR3_ANNOTATION_COST_PER_PAGE)
+
+
+def test_azure_doc_ai_annotation_pages_fall_back_to_ocr_rate(local_model_cost_map) -> None:
+    info = litellm.get_model_info(model=AZURE_DOC_AI_MODEL, custom_llm_provider="azure_ai")
+    assert info.get("annotation_cost_per_page") is None
+    assert info["ocr_cost_per_page"] == AZURE_DOC_AI_COST_PER_PAGE
+    cost = completion_cost(
+        completion_response=_annotated_ocr_response("mistral-document-ai-2512", 0, 1),
+        model=AZURE_DOC_AI_MODEL,
+        custom_llm_provider="azure_ai",
+        call_type="ocr",
+    )
+    assert cost == pytest.approx(AZURE_DOC_AI_COST_PER_PAGE)