import os import traceback from dotenv import load_dotenv load_dotenv() import json import tempfile from unittest.mock import ANY, AsyncMock, MagicMock, patch import httpx import pytest from respx import MockRouter import litellm from litellm import ( acompletion, completion, embedding, image_generation, ) from litellm.llms.vertex_ai.gemini.transformation import ( gemini_convert_messages_with_history, ) from litellm.llms.vertex_ai.vertex_llm_base import VertexBase litellm.num_retries = 3 litellm.cache = None user_message = "Write a short poem about the sky" messages = [{"content": user_message, "role": "user"}] VERTEX_MODELS_TO_NOT_TEST = [ "medlm-medium", "medlm-large", "code-gecko", "code-gecko@001", "code-gecko@002", "code-gecko@latest", "codechat-bison@latest", "code-bison@001", "text-bison@001", "gemini-1.5-pro", "gemini-1.5-pro-preview-0215", "gemini-pro-experimental", "gemini-flash-experimental", "gemini-2.5-flash-lite-exp-0827", "gemini-2.0-pro-exp-02-05", "gemini-pro-flash", "gemini-2.5-flash-lite-exp-0827", "gemini-2.0-flash-exp", "gemini-2.0-flash-thinking-exp", "gemini-2.0-flash-thinking-exp-01-21", "gemini-2.0-flash-preview-image-generation", "gemini-2.0-flash-live-preview-04-09", ] def load_vertex_ai_credentials(): # Define the path to the vertex_key.json file print("loading vertex ai credentials") filepath = os.path.dirname(os.path.abspath(__file__)) vertex_key_path = filepath + "/vertex_key.json" # Read the existing content of the file or create an empty dictionary try: with open(vertex_key_path, "r") as file: # Read the file content print("Read vertexai file path") content = file.read() # If the file is empty or not valid JSON, create an empty dictionary if not content or not content.strip(): service_account_key_data = {} else: # Attempt to load the existing JSON content file.seek(0) service_account_key_data = json.load(file) except FileNotFoundError: # If the file doesn't exist, create an empty dictionary service_account_key_data = {} # Update the service_account_key_data with environment variables private_key_id = os.environ.get("VERTEX_AI_PRIVATE_KEY_ID", "") private_key = os.environ.get("VERTEX_AI_PRIVATE_KEY", "") private_key = private_key.replace("\\n", "\n") service_account_key_data["private_key_id"] = private_key_id service_account_key_data["private_key"] = private_key # Create a temporary file with tempfile.NamedTemporaryFile(mode="w+", delete=False) as temp_file: # Write the updated content to the temporary files json.dump(service_account_key_data, temp_file, indent=2) # Export the temporary file as GOOGLE_APPLICATION_CREDENTIALS os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = os.path.abspath(temp_file.name) # test_vertex_ai_anthropic_streaming() # asyncio.run(test_vertex_ai_anthropic_async()) # asyncio.run(test_vertex_ai_anthropic_async_streaming()) # test_vertex_ai() # test_vertex_ai_stream() @pytest.mark.flaky(retries=3, delay=1) @pytest.mark.asyncio async def test_async_vertexai_streaming_response(): import random litellm.turn_on_debug() load_vertex_ai_credentials() test_models = ( litellm.vertex_chat_models | litellm.vertex_code_chat_models | litellm.vertex_text_models | litellm.vertex_code_text_models ) test_models = random.sample(list(test_models), 1) test_models += list(litellm.vertex_language_models) # always test gemini-pro test_models = ["gemini-3.5-flash"] for model in test_models: if model in VERTEX_MODELS_TO_NOT_TEST or ( "gecko" in model or "32k" in model or "ultra" in model or "002" in model or "gemini-2.0-flash-thinking-exp" in model or "gemini-2.0-pro-exp-02-05" in model or "gemini-pro" in model or "gemini-1.0-pro" in model or "image-generation" in model ): # our account does not have access to this model continue try: user_message = "Hello, how are you?" messages = [{"content": user_message, "role": "user"}] response = await acompletion( model=model, messages=messages, temperature=0.7, timeout=5, stream=True, vertex_location="global", ) print(f"response: {response}") complete_response: str = "" async for chunk in response: print(f"chunk: {chunk}") if chunk.choices[0].delta.content is not None: complete_response += chunk.choices[0].delta.content print(f"complete_response: {complete_response}") except litellm.NotFoundError as e: pass except litellm.RateLimitError as e: pass except litellm.APIConnectionError: pass except litellm.Timeout as e: pass except litellm.InternalServerError as e: pass except Exception as e: print(e) pytest.fail(f"An exception occurred: {e}") def vertex_httpx_grounding_post(*args, **kwargs): mock_response = MagicMock() mock_response.status_code = 200 mock_response.headers = {"Content-Type": "application/json"} mock_response.json.return_value = { "candidates": [ { "content": { "role": "model", "parts": [ { "text": "Argentina won the FIFA World Cup 2022. Argentina defeated France 4-2 on penalties in the FIFA World Cup 2022 final tournament for the first time after 36 years and the third time overall." } ], }, "finishReason": "STOP", "safetyRatings": [ { "category": "HARM_CATEGORY_HATE_SPEECH", "probability": "NEGLIGIBLE", "probabilityScore": 0.14940722, "severity": "HARM_SEVERITY_NEGLIGIBLE", "severityScore": 0.07477004, }, { "category": "HARM_CATEGORY_DANGEROUS_CONTENT", "probability": "NEGLIGIBLE", "probabilityScore": 0.15636235, "severity": "HARM_SEVERITY_NEGLIGIBLE", "severityScore": 0.015967654, }, { "category": "HARM_CATEGORY_HARASSMENT", "probability": "NEGLIGIBLE", "probabilityScore": 0.1943678, "severity": "HARM_SEVERITY_NEGLIGIBLE", "severityScore": 0.1284158, }, { "category": "HARM_CATEGORY_SEXUALLY_EXPLICIT", "probability": "NEGLIGIBLE", "probabilityScore": 0.09384396, "severity": "HARM_SEVERITY_NEGLIGIBLE", "severityScore": 0.0726367, }, ], "groundingMetadata": { "webSearchQueries": ["who won the world cup 2022"], "groundingAttributions": [ { "segment": {"endIndex": 38}, "confidenceScore": 0.9919262, "web": { "uri": "https://www.careerpower.in/fifa-world-cup-winners-list.html", "title": "FIFA World Cup Winners List from 1930 to 2022, Complete List - Career Power", }, }, { "segment": {"endIndex": 38}, "confidenceScore": 0.9919262, "web": { "uri": "https://www.careerpower.in/fifa-world-cup-winners-list.html", "title": "FIFA World Cup Winners List from 1930 to 2022, Complete List - Career Power", }, }, { "segment": {"endIndex": 38}, "confidenceScore": 0.9919262, "web": { "uri": "https://www.britannica.com/sports/2022-FIFA-World-Cup", "title": "2022 FIFA World Cup | Qatar, Controversy, Stadiums, Winner, & Final - Britannica", }, }, { "segment": {"endIndex": 38}, "confidenceScore": 0.9919262, "web": { "uri": "https://en.wikipedia.org/wiki/2022_FIFA_World_Cup_final", "title": "2022 FIFA World Cup final - Wikipedia", }, }, { "segment": {"endIndex": 38}, "confidenceScore": 0.9919262, "web": { "uri": "https://www.transfermarkt.com/2022-world-cup/erfolge/pokalwettbewerb/WM22", "title": "2022 World Cup - All winners - Transfermarkt", }, }, { "segment": {"startIndex": 39, "endIndex": 187}, "confidenceScore": 0.9919262, "web": { "uri": "https://www.careerpower.in/fifa-world-cup-winners-list.html", "title": "FIFA World Cup Winners List from 1930 to 2022, Complete List - Career Power", }, }, { "segment": {"startIndex": 39, "endIndex": 187}, "confidenceScore": 0.9919262, "web": { "uri": "https://en.wikipedia.org/wiki/2022_FIFA_World_Cup_final", "title": "2022 FIFA World Cup final - Wikipedia", }, }, ], "searchEntryPoint": { "renderedContent": '\u003cstyle\u003e\n.container {\n align-items: center;\n border-radius: 8px;\n display: flex;\n font-family: Google Sans, Roboto, sans-serif;\n font-size: 14px;\n line-height: 20px;\n padding: 8px 12px;\n}\n.chip {\n display: inline-block;\n border: solid 1px;\n border-radius: 16px;\n min-width: 14px;\n padding: 5px 16px;\n text-align: center;\n user-select: none;\n margin: 0 8px;\n -webkit-tap-highlight-color: transparent;\n}\n.carousel {\n overflow: auto;\n scrollbar-width: none;\n white-space: nowrap;\n margin-right: -12px;\n}\n.headline {\n display: flex;\n margin-right: 4px;\n}\n.gradient-container {\n position: relative;\n}\n.gradient {\n position: absolute;\n transform: translate(3px, -9px);\n height: 36px;\n width: 9px;\n}\n@media (prefers-color-scheme: light) {\n .container {\n background-color: #fafafa;\n box-shadow: 0 0 0 1px #0000000f;\n }\n .headline-label {\n color: #1f1f1f;\n }\n .chip {\n background-color: #ffffff;\n border-color: #d2d2d2;\n color: #5e5e5e;\n text-decoration: none;\n }\n .chip:hover {\n background-color: #f2f2f2;\n }\n .chip:focus {\n background-color: #f2f2f2;\n }\n .chip:active {\n background-color: #d8d8d8;\n border-color: #b6b6b6;\n }\n .logo-dark {\n display: none;\n }\n .gradient {\n background: linear-gradient(90deg, #fafafa 15%, #fafafa00 100%);\n }\n}\n@media (prefers-color-scheme: dark) {\n .container {\n background-color: #1f1f1f;\n box-shadow: 0 0 0 1px #ffffff26;\n }\n .headline-label {\n color: #fff;\n }\n .chip {\n background-color: #2c2c2c;\n border-color: #3c4043;\n color: #fff;\n text-decoration: none;\n }\n .chip:hover {\n background-color: #353536;\n }\n .chip:focus {\n background-color: #353536;\n }\n .chip:active {\n background-color: #464849;\n border-color: #53575b;\n }\n .logo-light {\n display: none;\n }\n .gradient {\n background: linear-gradient(90deg, #1f1f1f 15%, #1f1f1f00 100%);\n }\n}\n\u003c/style\u003e\n\u003cdiv class="container"\u003e\n \u003cdiv class="headline"\u003e\n \u003csvg class="logo-light" width="18" height="18" viewBox="9 9 35 35" fill="none" xmlns="http://www.w3.org/2000/svg"\u003e\n \u003cpath fill-rule="evenodd" clip-rule="evenodd" d="M42.8622 27.0064C42.8622 25.7839 42.7525 24.6084 42.5487 23.4799H26.3109V30.1568H35.5897C35.1821 32.3041 33.9596 34.1222 32.1258 35.3448V39.6864H37.7213C40.9814 36.677 42.8622 32.2571 42.8622 27.0064V27.0064Z" fill="#4285F4"/\u003e\n \u003cpath fill-rule="evenodd" clip-rule="evenodd" d="M26.3109 43.8555C30.9659 43.8555 34.8687 42.3195 37.7213 39.6863L32.1258 35.3447C30.5898 36.3792 28.6306 37.0061 26.3109 37.0061C21.8282 37.0061 18.0195 33.9811 16.6559 29.906H10.9194V34.3573C13.7563 39.9841 19.5712 43.8555 26.3109 43.8555V43.8555Z" fill="#34A853"/\u003e\n \u003cpath fill-rule="evenodd" clip-rule="evenodd" d="M16.6559 29.8904C16.3111 28.8559 16.1074 27.7588 16.1074 26.6146C16.1074 25.4704 16.3111 24.3733 16.6559 23.3388V18.8875H10.9194C9.74388 21.2072 9.06992 23.8247 9.06992 26.6146C9.06992 29.4045 9.74388 32.022 10.9194 34.3417L15.3864 30.8621L16.6559 29.8904V29.8904Z" fill="#FBBC05"/\u003e\n \u003cpath fill-rule="evenodd" clip-rule="evenodd" d="M26.3109 16.2386C28.85 16.2386 31.107 17.1164 32.9095 18.8091L37.8466 13.8719C34.853 11.082 30.9659 9.3736 26.3109 9.3736C19.5712 9.3736 13.7563 13.245 10.9194 18.8875L16.6559 23.3388C18.0195 19.2636 21.8282 16.2386 26.3109 16.2386V16.2386Z" fill="#EA4335"/\u003e\n \u003c/svg\u003e\n \u003csvg class="logo-dark" width="18" height="18" viewBox="0 0 48 48" xmlns="http://www.w3.org/2000/svg"\u003e\n \u003ccircle cx="24" cy="23" fill="#FFF" r="22"/\u003e\n \u003cpath d="M33.76 34.26c2.75-2.56 4.49-6.37 4.49-11.26 0-.89-.08-1.84-.29-3H24.01v5.99h8.03c-.4 2.02-1.5 3.56-3.07 4.56v.75l3.91 2.97h.88z" fill="#4285F4"/\u003e\n \u003cpath d="M15.58 25.77A8.845 8.845 0 0 0 24 31.86c1.92 0 3.62-.46 4.97-1.31l4.79 3.71C31.14 36.7 27.65 38 24 38c-5.93 0-11.01-3.4-13.45-8.36l.17-1.01 4.06-2.85h.8z" fill="#34A853"/\u003e\n \u003cpath d="M15.59 20.21a8.864 8.864 0 0 0 0 5.58l-5.03 3.86c-.98-2-1.53-4.25-1.53-6.64 0-2.39.55-4.64 1.53-6.64l1-.22 3.81 2.98.22 1.08z" fill="#FBBC05"/\u003e\n \u003cpath d="M24 14.14c2.11 0 4.02.75 5.52 1.98l4.36-4.36C31.22 9.43 27.81 8 24 8c-5.93 0-11.01 3.4-13.45 8.36l5.03 3.85A8.86 8.86 0 0 1 24 14.14z" fill="#EA4335"/\u003e\n \u003c/svg\u003e\n \u003cdiv class="gradient-container"\u003e\u003cdiv class="gradient"\u003e\u003c/div\u003e\u003c/div\u003e\n \u003c/div\u003e\n \u003cdiv class="carousel"\u003e\n \u003ca class="chip" href="https://www.google.com/search?q=who+won+the+world+cup+2022&client=app-vertex-grounding&safesearch=active"\u003ewho won the world cup 2022\u003c/a\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n' }, }, } ], "usageMetadata": { "promptTokenCount": 6, "candidatesTokenCount": 48, "totalTokenCount": 54, }, } return mock_response @pytest.mark.parametrize("value_in_dict", [{}, {"disable_attribution": False}]) # def test_gemini_pro_grounding(value_in_dict): try: load_vertex_ai_credentials() litellm.set_verbose = True tools = [{"googleSearchRetrieval": value_in_dict}] litellm.set_verbose = True from litellm.llms.custom_httpx.http_handler import HTTPHandler client = HTTPHandler() with patch.object( client, "post", side_effect=vertex_httpx_grounding_post ) as mock_call: resp = litellm.completion( model="vertex_ai_beta/gemini-1.0-pro-001", messages=[{"role": "user", "content": "Who won the world cup?"}], tools=tools, client=client, ) mock_call.assert_called_once() print(mock_call.call_args.kwargs["json"]["tools"][0]) assert ( "googleSearchRetrieval" in mock_call.call_args.kwargs["json"]["tools"][0] ) assert ( mock_call.call_args.kwargs["json"]["tools"][0]["googleSearchRetrieval"] == value_in_dict ) assert "vertex_ai_grounding_metadata" in resp._hidden_params assert isinstance(resp._hidden_params["vertex_ai_grounding_metadata"], list) except litellm.InternalServerError: pass except litellm.RateLimitError: pass from test_completion import response_format_tests @pytest.mark.parametrize( "model,region", [ ("vertex_ai/mistral-small-2503", "us-central1"), ("vertex_ai/qwen/qwen3-coder-480b-a35b-instruct-maas", "us-south1"), ("vertex_ai/openai/gpt-oss-20b-maas", "us-central1"), ], ) @pytest.mark.parametrize( "sync_mode", [True, False], ) # @pytest.mark.flaky(retries=3, delay=1) @pytest.mark.asyncio async def test_partner_models_httpx(model, region, sync_mode): try: load_vertex_ai_credentials() litellm.set_verbose = True messages = [ { "role": "system", "content": "Your name is Litellm Bot, you are a helpful assistant", }, # User asks for their name and weather in San Francisco { "role": "user", "content": "Hello, what is your name and can you tell me the weather?", }, ] data = { "model": model, "messages": messages, "timeout": 10, "vertex_ai_location": region, } if sync_mode: response = litellm.completion(**data) else: response = await litellm.acompletion(**data) response_format_tests(response=response) print(f"response: {response}") assert isinstance(response._hidden_params["response_cost"], float) except litellm.RateLimitError as e: print("RateLimitError", e) pass except litellm.Timeout as e: print("Timeout", e) pass except litellm.InternalServerError as e: print("InternalServerError", e) pass except litellm.APIConnectionError as e: print("APIConnectionError", e) pass except litellm.ServiceUnavailableError as e: print("ServiceUnavailableError", e) pass except Exception as e: print("got generic exception", e) if "429 Quota exceeded" in str(e): pass else: pytest.fail("An unexpected exception occurred - {}".format(str(e))) def vertex_httpx_mock_reject_prompt_post(*args, **kwargs): mock_response = MagicMock() mock_response.status_code = 200 mock_response.headers = {"Content-Type": "application/json"} mock_response.json.return_value = { "promptFeedback": {"blockReason": "OTHER"}, "usageMetadata": {"promptTokenCount": 6285, "totalTokenCount": 6285}, } return mock_response def vertex_httpx_mock_post(url, data=None, json=None, headers=None, **kwargs): mock_response = MagicMock() mock_response.status_code = 200 mock_response.headers = {"Content-Type": "application/json"} mock_response.json.return_value = { "candidates": [ { "finishReason": "RECITATION", "safetyRatings": [ { "category": "HARM_CATEGORY_HATE_SPEECH", "probability": "NEGLIGIBLE", "probabilityScore": 0.14965563, "severity": "HARM_SEVERITY_NEGLIGIBLE", "severityScore": 0.13660839, }, { "category": "HARM_CATEGORY_DANGEROUS_CONTENT", "probability": "NEGLIGIBLE", "probabilityScore": 0.16344544, "severity": "HARM_SEVERITY_NEGLIGIBLE", "severityScore": 0.10230471, }, { "category": "HARM_CATEGORY_HARASSMENT", "probability": "NEGLIGIBLE", "probabilityScore": 0.1979091, "severity": "HARM_SEVERITY_NEGLIGIBLE", "severityScore": 0.06052939, }, { "category": "HARM_CATEGORY_SEXUALLY_EXPLICIT", "probability": "NEGLIGIBLE", "probabilityScore": 0.1765296, "severity": "HARM_SEVERITY_NEGLIGIBLE", "severityScore": 0.18417984, }, ], "citationMetadata": { "citations": [ { "startIndex": 251, "endIndex": 380, "uri": "https://chocolatecake2023.blogspot.com/2023/02/taste-deliciousness-of-perfectly-baked.html?m=1", }, { "startIndex": 393, "endIndex": 535, "uri": "https://skinnymixes.co.uk/blogs/food-recipes/peanut-butter-cup-cookies", }, { "startIndex": 439, "endIndex": 581, "uri": "https://mast-producing-trees.org/aldis-chocolate-chips-are-peanut-and-tree-nut-free/", }, { "startIndex": 1117, "endIndex": 1265, "uri": "https://github.com/frdrck100/To_Do_Assignments", }, { "startIndex": 1146, "endIndex": 1288, "uri": "https://skinnymixes.co.uk/blogs/food-recipes/peanut-butter-cup-cookies", }, { "startIndex": 1166, "endIndex": 1299, "uri": "https://www.girlversusdough.com/brookies/", }, { "startIndex": 1780, "endIndex": 1909, "uri": "https://chocolatecake2023.blogspot.com/2023/02/taste-deliciousness-of-perfectly-baked.html?m=1", }, { "startIndex": 1834, "endIndex": 1964, "uri": "https://newsd.in/national-cream-cheese-brownie-day-2023-date-history-how-to-make-a-cream-cheese-brownie/", }, { "startIndex": 1846, "endIndex": 1989, "uri": "https://github.com/frdrck100/To_Do_Assignments", }, { "startIndex": 2121, "endIndex": 2261, "uri": "https://recipes.net/copycat/hardee/hardees-chocolate-chip-cookie-recipe/", }, { "startIndex": 2505, "endIndex": 2671, "uri": "https://www.tfrecipes.com/Oranges%20with%20dried%20cherries/", }, { "startIndex": 3390, "endIndex": 3529, "uri": "https://github.com/quantumcognition/Crud-palm", }, { "startIndex": 3568, "endIndex": 3724, "uri": "https://recipes.net/dessert/cakes/ultimate-easy-gingerbread/", }, { "startIndex": 3640, "endIndex": 3770, "uri": "https://recipes.net/dessert/cookies/soft-and-chewy-peanut-butter-cookies/", }, ] }, } ], "usageMetadata": {"promptTokenCount": 336, "totalTokenCount": 336}, } return mock_response @pytest.mark.parametrize("provider", ["vertex_ai_beta"]) # "vertex_ai", @pytest.mark.parametrize("content_filter_type", ["prompt", "response"]) # "vertex_ai", @pytest.mark.asyncio @pytest.mark.flaky(retries=3, delay=1) async def test_gemini_pro_json_schema_httpx_content_policy_error( provider, content_filter_type ): load_vertex_ai_credentials() litellm.set_verbose = True messages = [ { "role": "user", "content": """ List 5 popular cookie recipes. Using this JSON schema: ```json {'$defs': {'Recipe': {'properties': {'recipe_name': {'examples': ['Chocolate Chip Cookies', 'Peanut Butter Cookies'], 'maxLength': 100, 'title': 'The recipe name', 'type': 'string'}, 'estimated_time': {'anyOf': [{'minimum': 0, 'type': 'integer'}, {'type': 'null'}], 'default': None, 'description': 'The estimated time to make the recipe in minutes', 'examples': [30, 45], 'title': 'The estimated time'}, 'ingredients': {'examples': [['flour', 'sugar', 'chocolate chips'], ['peanut butter', 'sugar', 'eggs']], 'items': {'type': 'string'}, 'maxItems': 10, 'title': 'The ingredients', 'type': 'array'}, 'instructions': {'examples': [['mix', 'bake'], ['mix', 'chill', 'bake']], 'items': {'type': 'string'}, 'maxItems': 10, 'title': 'The instructions', 'type': 'array'}}, 'required': ['recipe_name', 'ingredients', 'instructions'], 'title': 'Recipe', 'type': 'object'}}, 'properties': {'recipes': {'items': {'$ref': '#/$defs/Recipe'}, 'maxItems': 11, 'title': 'The recipes', 'type': 'array'}}, 'required': ['recipes'], 'title': 'MyRecipes', 'type': 'object'} ``` """, } ] from litellm.llms.custom_httpx.http_handler import HTTPHandler client = HTTPHandler() if content_filter_type == "prompt": _side_effect = vertex_httpx_mock_reject_prompt_post else: _side_effect = vertex_httpx_mock_post with patch.object(client, "post", side_effect=_side_effect) as mock_call: response = completion( model="vertex_ai_beta/gemini-2.5-flash-lite", messages=messages, response_format={"type": "json_object"}, client=client, logging_obj=ANY, ) assert response.choices[0].finish_reason == "content_filter" mock_call.assert_called_once() def vertex_httpx_mock_post_valid_response(*args, **kwargs): mock_response = MagicMock() mock_response.status_code = 200 mock_response.headers = {"Content-Type": "application/json"} mock_response.json.return_value = { "candidates": [ { "content": { "role": "model", "parts": [ { "text": """{ "recipes": [ {"recipe_name": "Chocolate Chip Cookies"}, {"recipe_name": "Oatmeal Raisin Cookies"}, {"recipe_name": "Peanut Butter Cookies"}, {"recipe_name": "Sugar Cookies"}, {"recipe_name": "Snickerdoodles"} ] }""" } ], }, "finishReason": "STOP", "safetyRatings": [ { "category": "HARM_CATEGORY_HATE_SPEECH", "probability": "NEGLIGIBLE", "probabilityScore": 0.09790669, "severity": "HARM_SEVERITY_NEGLIGIBLE", "severityScore": 0.11736965, }, { "category": "HARM_CATEGORY_DANGEROUS_CONTENT", "probability": "NEGLIGIBLE", "probabilityScore": 0.1261379, "severity": "HARM_SEVERITY_NEGLIGIBLE", "severityScore": 0.08601588, }, { "category": "HARM_CATEGORY_HARASSMENT", "probability": "NEGLIGIBLE", "probabilityScore": 0.083441176, "severity": "HARM_SEVERITY_NEGLIGIBLE", "severityScore": 0.0355444, }, { "category": "HARM_CATEGORY_SEXUALLY_EXPLICIT", "probability": "NEGLIGIBLE", "probabilityScore": 0.071981624, "severity": "HARM_SEVERITY_NEGLIGIBLE", "severityScore": 0.08108212, }, ], } ], "usageMetadata": { "promptTokenCount": 60, "candidatesTokenCount": 55, "totalTokenCount": 115, }, } return mock_response def vertex_httpx_mock_post_valid_response_anthropic(*args, **kwargs): mock_response = MagicMock() mock_response.status_code = 200 mock_response.headers = {"Content-Type": "application/json"} mock_response.json.return_value = { "id": "msg_vrtx_013Wki5RFQXAspL7rmxRFjZg", "type": "message", "role": "assistant", "model": "claude-3-5-sonnet-20240620", "content": [ { "type": "tool_use", "id": "toolu_vrtx_01YMnYZrToPPfcmY2myP2gEB", "name": "json_tool_call", "input": { "values": { "recipes": [ {"recipe_name": "Chocolate Chip Cookies"}, {"recipe_name": "Oatmeal Raisin Cookies"}, {"recipe_name": "Peanut Butter Cookies"}, {"recipe_name": "Snickerdoodle Cookies"}, {"recipe_name": "Sugar Cookies"}, ] } }, } ], "stop_reason": "tool_use", "stop_sequence": None, "usage": {"input_tokens": 368, "output_tokens": 118}, } return mock_response def vertex_httpx_mock_post_invalid_schema_response(*args, **kwargs): mock_response = MagicMock() mock_response.status_code = 200 mock_response.headers = {"Content-Type": "application/json"} mock_response.json.return_value = { "candidates": [ { "content": { "role": "model", "parts": [ {"text": '[{"recipe_world": "Chocolate Chip Cookies"}]\n'} ], }, "finishReason": "STOP", "safetyRatings": [ { "category": "HARM_CATEGORY_HATE_SPEECH", "probability": "NEGLIGIBLE", "probabilityScore": 0.09790669, "severity": "HARM_SEVERITY_NEGLIGIBLE", "severityScore": 0.11736965, }, { "category": "HARM_CATEGORY_DANGEROUS_CONTENT", "probability": "NEGLIGIBLE", "probabilityScore": 0.1261379, "severity": "HARM_SEVERITY_NEGLIGIBLE", "severityScore": 0.08601588, }, { "category": "HARM_CATEGORY_HARASSMENT", "probability": "NEGLIGIBLE", "probabilityScore": 0.083441176, "severity": "HARM_SEVERITY_NEGLIGIBLE", "severityScore": 0.0355444, }, { "category": "HARM_CATEGORY_SEXUALLY_EXPLICIT", "probability": "NEGLIGIBLE", "probabilityScore": 0.071981624, "severity": "HARM_SEVERITY_NEGLIGIBLE", "severityScore": 0.08108212, }, ], } ], "usageMetadata": { "promptTokenCount": 60, "candidatesTokenCount": 55, "totalTokenCount": 115, }, } return mock_response def vertex_httpx_mock_post_invalid_schema_response_anthropic(*args, **kwargs): mock_response = MagicMock() mock_response.status_code = 200 mock_response.headers = {"Content-Type": "application/json"} mock_response.json.return_value = { "id": "msg_vrtx_013Wki5RFQXAspL7rmxRFjZg", "type": "message", "role": "assistant", "model": "claude-3-5-sonnet-20240620", "content": [{"text": "Hi! My name is Claude.", "type": "text"}], "stop_reason": "end_turn", "stop_sequence": None, "usage": {"input_tokens": 368, "output_tokens": 118}, } return mock_response @pytest.mark.parametrize( "model, vertex_location, supports_response_schema", [ ("vertex_ai_beta/gemini-2.0-flash-001", "us-central1", True), ("vertex_ai_beta/gemini-2.5-flash-lite", "us-central1", True), ("vertex_ai/claude-3-5-sonnet@20240620", "us-east5", False), ], ) @pytest.mark.parametrize("invalid_response", [True, False]) @pytest.mark.parametrize("enforce_validation", [True, False]) @pytest.mark.asyncio async def test_gemini_pro_json_schema_args_sent_httpx( model, supports_response_schema, vertex_location, invalid_response, enforce_validation, ): load_vertex_ai_credentials() os.environ["LITELLM_LOCAL_MODEL_COST_MAP"] = "True" litellm.model_cost = litellm.get_model_cost_map(url="") litellm.set_verbose = True messages = [{"role": "user", "content": "List 5 cookie recipes"}] from litellm.llms.custom_httpx.http_handler import HTTPHandler response_schema = { "type": "object", "properties": { "recipes": { "type": "array", "items": { "type": "object", "properties": {"recipe_name": {"type": "string"}}, "required": ["recipe_name"], }, } }, "required": ["recipes"], "additionalProperties": False, } client = HTTPHandler() httpx_response = MagicMock() if invalid_response is True: if "claude" in model: httpx_response.side_effect = vertex_httpx_mock_post_invalid_schema_response_anthropic else: httpx_response.side_effect = vertex_httpx_mock_post_invalid_schema_response else: if "claude" in model: httpx_response.side_effect = vertex_httpx_mock_post_valid_response_anthropic else: httpx_response.side_effect = vertex_httpx_mock_post_valid_response resp = None with patch.object(client, "post", new=httpx_response) as mock_call: litellm.set_verbose = True print(f"model entering completion: {model}") try: resp = completion( model=model, messages=messages, response_format={ "type": "json_object", "response_schema": response_schema, "enforce_validation": enforce_validation, }, vertex_location=vertex_location, client=client, ) print("Received={}".format(resp)) if invalid_response is True and enforce_validation is True: pytest.fail("Expected this to fail") except litellm.JSONSchemaValidationError as e: if invalid_response is False: pytest.fail("Expected this to pass. Got={}".format(e)) mock_call.assert_called_once() if "claude" not in model: print(mock_call.call_args.kwargs) print(mock_call.call_args.kwargs["json"]["generationConfig"]) if supports_response_schema: gen_config = mock_call.call_args.kwargs["json"]["generationConfig"] assert ( "response_schema" in gen_config or "response_json_schema" in gen_config ), f"Expected response_schema or response_json_schema in {gen_config}" else: gen_config = mock_call.call_args.kwargs["json"]["generationConfig"] assert ( "response_schema" not in gen_config and "response_json_schema" not in gen_config ) assert ( "Use this JSON schema:" in mock_call.call_args.kwargs["json"]["contents"][0]["parts"][1]["text"] ) elif resp is not None: assert resp.model == model.split("/")[1] @pytest.mark.asyncio async def test_anthropic_message_via_anthropic_messages(): from unittest.mock import AsyncMock from litellm.llms.custom_httpx.llm_http_handler import AsyncHTTPHandler load_vertex_ai_credentials() os.environ["LITELLM_LOCAL_MODEL_COST_MAP"] = "True" litellm.model_cost = litellm.get_model_cost_map(url="") litellm.set_verbose = True client = AsyncHTTPHandler() httpx_response = AsyncMock() httpx_response.side_effect = vertex_httpx_mock_post_valid_response_anthropic call_1_kwargs = {} call_2_kwargs = {} with patch.object(client, "post", new=httpx_response) as mock_call: messages = [{"role": "user", "content": "List 5 cookie recipes"}] response = await litellm.anthropic_messages( model="vertex_ai/claude-3-5-sonnet@20240620", messages=messages, max_tokens=100, client=client, ) print(f"response: {response}") assert mock_call.call_count == 1 call_1_kwargs = mock_call.call_args.kwargs with patch.object(client, "post", new=httpx_response) as mock_call: response_2 = await litellm.acompletion( model="vertex_ai/claude-3-5-sonnet@20240620", messages=messages, max_tokens=100, client=client, ) print(f"response_2: {response_2}") call_args = mock_call.call_args print(f"call_args: {call_args}") call_2_kwargs = mock_call.call_args.kwargs call_2_kwargs["url"] = call_args[0][0] """ Compare Call 1 and Call 2 Expect: - url - headers - data / json to be the same, except for the Authorization header. """ print(f"call_1_kwargs: {call_1_kwargs}") print(f"call_2_kwargs: {call_2_kwargs}") assert ( call_1_kwargs["url"] == call_2_kwargs["url"] ), f"Expected url to be the same, but got {call_1_kwargs['url']} and Expected {call_2_kwargs['url']}" assert "Authorization".lower() in [ k.lower() for k in call_1_kwargs["headers"].keys() ], f"Expected Authorization header to be present in call_1_kwargs, but got {call_1_kwargs['headers'].keys()}" assert "content-type".lower() in [ k.lower() for k in call_1_kwargs["headers"].keys() ], f"Expected Content-Type header to be present in call_1_kwargs, but got {call_1_kwargs['headers'].keys()}" ## validate request body print(f"call 1 kwargs keys: {call_1_kwargs.keys()}") print(f"call_2_kwargs['json']: {type(call_2_kwargs['json'])}") print(f"call_1_kwargs['data']: {type(call_1_kwargs['data'])}") call_1_kwargs_data = json.loads(call_1_kwargs["data"]) for k, v in call_2_kwargs["json"].items(): assert ( k in call_1_kwargs_data ), f"Expected {k} to be present in call_1_kwargs['data'], but got {call_1_kwargs_data.keys()}" @pytest.mark.parametrize( "model, vertex_location, supports_response_schema", [ ("vertex_ai_beta/gemini-2.0-flash-001", "us-central1", True), ("vertex_ai_beta/gemini-2.5-flash-lite", "us-central1", True), ("vertex_ai/claude-3-5-sonnet@20240620", "us-east5", False), ], ) @pytest.mark.parametrize("invalid_response", [True, False]) @pytest.mark.parametrize("enforce_validation", [True, False]) @pytest.mark.asyncio async def test_gemini_pro_json_schema_args_sent_httpx_openai_schema( model, supports_response_schema, vertex_location, invalid_response, enforce_validation, ): from typing import List if enforce_validation: litellm.enable_json_schema_validation = True from pydantic import BaseModel load_vertex_ai_credentials() os.environ["LITELLM_LOCAL_MODEL_COST_MAP"] = "True" litellm.model_cost = litellm.get_model_cost_map(url="") litellm.set_verbose = True messages = [{"role": "user", "content": "List 5 cookie recipes"}] from litellm.llms.custom_httpx.http_handler import HTTPHandler class Recipe(BaseModel): recipe_name: str class ResponseSchema(BaseModel): recipes: List[Recipe] client = HTTPHandler() httpx_response = MagicMock() if invalid_response is True: if "claude" in model: httpx_response.side_effect = vertex_httpx_mock_post_invalid_schema_response_anthropic else: httpx_response.side_effect = vertex_httpx_mock_post_invalid_schema_response else: if "claude" in model: httpx_response.side_effect = vertex_httpx_mock_post_valid_response_anthropic else: httpx_response.side_effect = vertex_httpx_mock_post_valid_response with patch.object(client, "post", new=httpx_response) as mock_call: print("SENDING CLIENT POST={}".format(client.post)) try: resp = completion( model=model, messages=messages, response_format=ResponseSchema, vertex_location=vertex_location, client=client, ) print("Received={}".format(resp)) if invalid_response is True and enforce_validation is True: pytest.fail("Expected this to fail") except litellm.JSONSchemaValidationError as e: if invalid_response is False: pytest.fail("Expected this to pass. Got={}".format(e)) mock_call.assert_called_once() if "claude" not in model: print(mock_call.call_args.kwargs) print(mock_call.call_args.kwargs["json"]["generationConfig"]) if supports_response_schema: gen_config = mock_call.call_args.kwargs["json"]["generationConfig"] assert ( "response_schema" in gen_config or "response_json_schema" in gen_config ), f"Expected response_schema or response_json_schema in {gen_config}" assert ( "response_mime_type" in mock_call.call_args.kwargs["json"]["generationConfig"] ) assert ( mock_call.call_args.kwargs["json"]["generationConfig"]["response_mime_type"] == "application/json" ) else: gen_config = mock_call.call_args.kwargs["json"]["generationConfig"] assert ( "response_schema" not in gen_config and "response_json_schema" not in gen_config ) assert ( "Use this JSON schema:" in mock_call.call_args.kwargs["json"]["contents"][0]["parts"][1]["text"] ) @pytest.mark.parametrize( "model", ["gemini-2.5-flash-lite", "claude-3-5-sonnet@20240620"] ) @pytest.mark.asyncio async def test_gemini_pro_httpx_custom_api_base(model): load_vertex_ai_credentials() litellm.set_verbose = True messages = [ { "role": "user", "content": "Hello world", } ] from litellm.llms.custom_httpx.http_handler import HTTPHandler client = HTTPHandler() with patch.object(client, "post", new=MagicMock()) as mock_call: try: response = completion( model="vertex_ai/{}".format(model), messages=messages, response_format={"type": "json_object"}, client=client, api_base="my-custom-api-base", extra_headers={"hello": "world"}, ) except Exception as e: traceback.print_exc() print("Receives error - {}".format(str(e))) mock_call.assert_called_once() print(f"mock_call.call_args: {mock_call.call_args}") print(f"mock_call.call_args.kwargs: {mock_call.call_args.kwargs}") if "url" in mock_call.call_args.kwargs: assert ( "my-custom-api-base:generateContent" == mock_call.call_args.kwargs["url"] ) else: assert "my-custom-api-base:rawPredict" == mock_call.call_args[0][0] if "headers" in mock_call.call_args.kwargs: assert "hello" in mock_call.call_args.kwargs["headers"] # gemini_pro_function_calling() # asyncio.run(gemini_pro_async_function_calling()) @pytest.mark.parametrize( ("route", "provider"), [ ("completion", "vertex_ai"), ("embedding", "vertex_ai"), ("image_generation", "gemini"), ], ids=[ "completion-vertex_ai", "embedding-vertex_ai", "image_generation-gemini", ], ) def test_litellm_api_base(monkeypatch, route, provider): from litellm.llms.custom_httpx.http_handler import HTTPHandler client = HTTPHandler() import litellm monkeypatch.setattr(litellm, "api_base", "https://litellm.com") load_vertex_ai_credentials() if route == "image_generation" and provider == "gemini": pytest.skip("Gemini does not support image generation") with patch.object(client, "post", new=MagicMock()) as mock_client: try: if route == "completion": response = completion( model=f"{provider}/gemini-2.0-flash-001", messages=[{"role": "user", "content": "Hello, world!"}], client=client, ) elif route == "embedding": response = embedding( model=f"{provider}/gemini-2.0-flash-001", input=["Hello, world!"], client=client, ) elif route == "image_generation": response = image_generation( model=f"{provider}/gemini-2.0-flash-001", prompt="Hello, world!", client=client, ) except Exception as e: print(e) mock_client.assert_called() assert mock_client.call_args.kwargs["url"].startswith("https://litellm.com") def test_prompt_factory(): messages = [ { "role": "system", "content": "Your name is Litellm Bot, you are a helpful assistant", }, # User asks for their name and weather in San Francisco { "role": "user", "content": "Hello, what is your name and can you tell me the weather?", }, # Assistant replies with a tool call { "role": "assistant", "content": "", "tool_calls": [ { "id": "call_123", "type": "function", "index": 0, "function": { "name": "get_weather", "arguments": '{"location":"San Francisco, CA"}', }, } ], }, # The result of the tool call is added to the history { "role": "tool", "tool_call_id": "call_123", "content": "27 degrees celsius and clear in San Francisco, CA", }, # Now the assistant can reply with the result of the tool call. ] translated_messages = gemini_convert_messages_with_history(messages=messages) print(f"\n\ntranslated_messages: {translated_messages}\ntranslated_messages") @pytest.mark.asyncio async def test_completion_fine_tuned_model(): load_vertex_ai_credentials() mock_response = AsyncMock() mock_response.headers = {} mock_response.status_code = 200 def return_val(): return { "candidates": [ { "content": { "role": "model", "parts": [ { "text": "A canvas vast, a boundless blue,\nWhere clouds paint tales and winds imbue.\nThe sun descends in fiery hue,\nStars shimmer bright, a gentle few.\n\nThe moon ascends, a pearl of light,\nGuiding travelers through the night.\nThe sky embraces, holds all tight,\nA tapestry of wonder, bright." } ], }, "finishReason": "STOP", "safetyRatings": [ { "category": "HARM_CATEGORY_HATE_SPEECH", "probability": "NEGLIGIBLE", "probabilityScore": 0.028930664, "severity": "HARM_SEVERITY_NEGLIGIBLE", "severityScore": 0.041992188, }, # ... other safety ratings ... ], "avgLogprobs": -0.95772853367765187, } ], "usageMetadata": { "promptTokenCount": 7, "candidatesTokenCount": 71, "totalTokenCount": 78, }, } mock_response.json = return_val expected_payload = { "contents": [ {"role": "user", "parts": [{"text": "Write a short poem about the sky"}]} ] } with patch( "litellm.llms.custom_httpx.http_handler.AsyncHTTPHandler.post", return_value=mock_response, ) as mock_post: # Act: Call the litellm.completion function response = await litellm.acompletion( model="vertex_ai_beta/4965075652664360960", messages=[{"role": "user", "content": "Write a short poem about the sky"}], ) # Assert mock_post.assert_called_once() url, kwargs = mock_post.call_args print("url = ", url) # this is the fine-tuned model endpoint assert ( url[0] == "https://us-central1-aiplatform.googleapis.com/v1/projects/litellm-ci-cd/locations/us-central1/endpoints/4965075652664360960:generateContent" ) print("call args = ", kwargs) args_to_vertexai = kwargs["json"] print("args to vertex ai call:", args_to_vertexai) assert args_to_vertexai == expected_payload assert response.choices[0].message.content.startswith("A canvas vast") assert response.choices[0].finish_reason == "stop" assert response.usage.total_tokens == 78 # Optional: Print for debugging print("Arguments passed to Vertex AI:", args_to_vertexai) print("Response:", response) def mock_gemini_request(*args, **kwargs): print(f"kwargs: {kwargs}") mock_response = MagicMock() mock_response.status_code = 200 mock_response.headers = {"Content-Type": "application/json"} if "cachedContents" in kwargs["url"]: mock_response.json.return_value = { "name": "cachedContents/4d2kd477o3pg", "model": "models/gemini-2.5-flash-lite-001", "createTime": "2024-08-26T22:31:16.147190Z", "updateTime": "2024-08-26T22:31:16.147190Z", "expireTime": "2024-08-26T22:36:15.548934784Z", "displayName": "", "usageMetadata": {"totalTokenCount": 323383}, } else: mock_response.json.return_value = { "candidates": [ { "content": { "parts": [ { "text": "Please provide me with the text of the legal agreement" } ], "role": "model", }, "finishReason": "MAX_TOKENS", "index": 0, "safetyRatings": [ { "category": "HARM_CATEGORY_SEXUALLY_EXPLICIT", "probability": "NEGLIGIBLE", }, { "category": "HARM_CATEGORY_HATE_SPEECH", "probability": "NEGLIGIBLE", }, { "category": "HARM_CATEGORY_HARASSMENT", "probability": "NEGLIGIBLE", }, { "category": "HARM_CATEGORY_DANGEROUS_CONTENT", "probability": "NEGLIGIBLE", }, ], } ], "usageMetadata": { "promptTokenCount": 40049, "candidatesTokenCount": 10, "totalTokenCount": 40059, "cachedContentTokenCount": 40012, }, } return mock_response def mock_gemini_list_request(*args, **kwargs): from litellm.types.llms.vertex_ai import ( CachedContent, CachedContentListAllResponseBody, ) print(f"kwargs: {kwargs}") mock_response = MagicMock() mock_response.status_code = 200 mock_response.headers = {"Content-Type": "application/json"} mock_response.json.return_value = CachedContentListAllResponseBody( cachedContents=[CachedContent(name="test", displayName="test")] ) return mock_response from litellm._uuid import uuid @pytest.mark.parametrize( "sync_mode", [True, False], ) @pytest.mark.asyncio async def test_gemini_context_caching_anthropic_format(sync_mode): from litellm.llms.custom_httpx.http_handler import AsyncHTTPHandler, HTTPHandler litellm.set_verbose = True gemini_context_caching_messages = [ # System Message { "role": "system", "content": [ { "type": "text", "text": "Here is the full text of a complex legal agreement {}".format( uuid.uuid4() ) * 4000, "cache_control": {"type": "ephemeral"}, } ], }, # marked for caching with the cache_control parameter, so that this checkpoint can read from the previous cache. { "role": "user", "content": [ { "type": "text", "text": "What are the key terms and conditions in this agreement?", "cache_control": {"type": "ephemeral"}, } ], }, { "role": "assistant", "content": "Certainly! the key terms and conditions are the following: the contract is 1 year long for $10/mo", }, # The final turn is marked with cache-control, for continuing in followups. { "role": "user", "content": [ { "type": "text", "text": "What are the key terms and conditions in this agreement?", } ], }, ] if sync_mode: client = HTTPHandler(concurrent_limit=1) else: client = AsyncHTTPHandler(concurrent_limit=1) with patch.object(client, "post", side_effect=mock_gemini_request) as mock_client: try: if sync_mode: response = litellm.completion( model="gemini/gemini-2.5-flash-lite-001", messages=gemini_context_caching_messages, temperature=0.2, max_tokens=10, client=client, ) else: response = await litellm.acompletion( model="gemini/gemini-2.5-flash-lite-001", messages=gemini_context_caching_messages, temperature=0.2, max_tokens=10, client=client, ) except Exception as e: print(e) assert mock_client.call_count == 2 first_call_args = mock_client.call_args_list[0].kwargs print(f"first_call_args: {first_call_args}") assert "cachedContents" in first_call_args["url"] # assert "cache_read_input_tokens" in response.usage # assert "cache_creation_input_tokens" in response.usage # # Assert either a cache entry was created or cache was read - changes depending on the anthropic api ttl # assert (response.usage.cache_read_input_tokens > 0) or ( # response.usage.cache_creation_input_tokens > 0 # ) @pytest.mark.asyncio async def test_partner_models_httpx_ai21(): litellm.set_verbose = True model = "vertex_ai/jamba-1.5-mini@001" messages = [ { "role": "system", "content": "Your name is Litellm Bot, you are a helpful assistant", }, { "role": "user", "content": "Hello, can you tell me the weather in San Francisco?", }, ] tools = [ { "type": "function", "function": { "name": "get_weather", "description": "Get the current weather in a given location", "parameters": { "type": "object", "properties": { "location": { "type": "string", "description": "The city and state, e.g. San Francisco, CA", } }, "required": ["location"], }, }, } ] data = { "model": model, "messages": messages, "tools": tools, "top_p": 0.5, } mock_response = AsyncMock() def return_val(): return { "id": "chat-3d11cf95eb224966937b216d9494fe73", "choices": [ { "index": 0, "message": { "role": "assistant", "content": " Sure, let me check that for you.", "tool_calls": [ { "id": "b5cef16b-5946-4937-b9d5-beeaea871e77", "type": "function", "function": { "name": "get_weather", "arguments": '{"location": "San Francisco"}', }, } ], }, "finish_reason": "stop", } ], "usage": { "prompt_tokens": 158, "completion_tokens": 36, "total_tokens": 194, }, "meta": {"requestDurationMillis": 501}, "model": "jamba-1.5-mini@001", } mock_response.json = return_val mock_response.status_code = 200 with patch( "litellm.llms.custom_httpx.http_handler.AsyncHTTPHandler.post", return_value=mock_response, ) as mock_post: response = await litellm.acompletion(**data) # Assert mock_post.assert_called_once() url, kwargs = mock_post.call_args print("url = ", url) print("call args = ", kwargs) print(kwargs["data"]) assert ( url[0] == "https://us-central1-aiplatform.googleapis.com/v1beta1/projects/litellm-ci-cd/locations/us-central1/publishers/ai21/models/jamba-1.5-mini@001:rawPredict" ) # json loads kwargs kwargs["data"] = json.loads(kwargs["data"]) assert kwargs["data"] == { "model": "jamba-1.5-mini@001", "messages": [ { "role": "system", "content": "Your name is Litellm Bot, you are a helpful assistant", }, { "role": "user", "content": "Hello, can you tell me the weather in San Francisco?", }, ], "top_p": 0.5, "tools": [ { "type": "function", "function": { "name": "get_weather", "description": "Get the current weather in a given location", "parameters": { "type": "object", "properties": { "location": { "type": "string", "description": "The city and state, e.g. San Francisco, CA", } }, "required": ["location"], }, }, } ], "stream": False, } assert response.id == "chat-3d11cf95eb224966937b216d9494fe73" assert len(response.choices) == 1 assert ( response.choices[0].message.content == " Sure, let me check that for you." ) assert response.choices[0].message.tool_calls[0].function.name == "get_weather" assert ( response.choices[0].message.tool_calls[0].function.arguments == '{"location": "San Francisco"}' ) assert response.usage.prompt_tokens == 158 assert response.usage.completion_tokens == 36 assert response.usage.total_tokens == 194 print(f"response: {response}") @pytest.mark.parametrize( "base_model, metadata", [ (None, {"model_info": {"base_model": "vertex_ai/gemini-1.5-pro"}}), ("vertex_ai/gemini-1.5-pro", None), ], ) def test_gemini_finetuned_endpoint(base_model, metadata): litellm.set_verbose = True load_vertex_ai_credentials() from litellm.llms.custom_httpx.http_handler import HTTPHandler # Set up the messages messages = [ {"role": "system", "content": """Use search for most queries."""}, {"role": "user", "content": """search for weather in boston (use `search`)"""}, ] client = HTTPHandler(concurrent_limit=1) with patch.object(client, "post", new=MagicMock()) as mock_client: try: response = completion( model="vertex_ai/4965075652664360960", messages=messages, tool_choice="auto", client=client, metadata=metadata, base_model=base_model, ) except Exception as e: print(e) print(mock_client.call_args.kwargs) mock_client.assert_called() assert mock_client.call_args.kwargs["url"].endswith( "endpoints/4965075652664360960:generateContent" ) @pytest.mark.asyncio @pytest.mark.respx async def test_vertexai_embedding_finetuned(respx_mock: MockRouter): """ Tests that: - Request URL and body are correctly formatted for Vertex AI embeddings - Response is properly parsed into litellm's embedding response format """ load_vertex_ai_credentials() litellm.set_verbose = True litellm.disable_aiohttp_transport = ( True # since this uses respx, we need to set use_aiohttp_transport to False ) # Test input input_text = ["good morning from litellm", "this is another item"] # Expected request/response expected_url = "https://us-central1-aiplatform.googleapis.com/v1/projects/633608382793/locations/us-central1/endpoints/1004708436694269952:predict" expected_request = { "instances": [ {"inputs": "good morning from litellm"}, {"inputs": "this is another item"}, ], "parameters": {}, } mock_response = { "predictions": [ [[-0.000431762, -0.04416759, -0.03443353]], # Truncated embedding vector [[-0.000431762, -0.04416759, -0.03443353]], # Truncated embedding vector ], "deployedModelId": "2275167734310371328", "model": "projects/633608382793/locations/us-central1/models/snowflake-arctic-embed-m-long-1731622468876", "modelDisplayName": "snowflake-arctic-embed-m-long-1731622468876", "modelVersionId": "1", } # Setup mock request mock_request = respx_mock.post(expected_url).mock( return_value=httpx.Response(200, json=mock_response) ) # Make request response = await litellm.aembedding( vertex_project="633608382793", model="vertex_ai/1004708436694269952", input=input_text, ) # Assert request was made correctly assert mock_request.called request_body = json.loads(mock_request.calls[0].request.content) print("\n\nrequest_body", request_body) print("\n\nexpected_request", expected_request) assert request_body == expected_request # Assert response structure assert response is not None assert hasattr(response, "data") assert len(response.data) == len(input_text) # Assert embedding structure for embedding in response.data: assert "embedding" in embedding assert isinstance(embedding["embedding"], list) assert len(embedding["embedding"]) > 0 assert all(isinstance(x, float) for x in embedding["embedding"]) @pytest.mark.parametrize("max_retries", [None, 3]) @pytest.mark.asyncio @pytest.mark.respx async def test_vertexai_model_garden_model_completion( respx_mock: MockRouter, max_retries ): """ Relevant issue: https://github.com/BerriAI/litellm/issues/6480 Using OpenAI compatible models from Vertex Model Garden """ litellm.disable_aiohttp_transport = ( True # since this uses respx, we need to set use_aiohttp_transport to False ) litellm.module_level_aclient = httpx.AsyncClient() load_vertex_ai_credentials() litellm.set_verbose = True # Test input messages = [ { "role": "system", "content": "Your name is Litellm Bot, you are a helpful assistant", }, { "role": "user", "content": "Hello, what is your name and can you tell me the weather?", }, ] # Expected request/response expected_url = "https://us-central1-aiplatform.googleapis.com/v1beta1/projects/633608382793/locations/us-central1/endpoints/5464397967697903616/chat/completions" expected_request = {"model": "", "messages": messages, "stream": False} mock_response = { "id": "chat-09940d4e99e3488aa52a6f5e2ecf35b1", "object": "chat.completion", "created": 1731702782, "model": "meta-llama/Llama-3.1-8B-Instruct", "choices": [ { "index": 0, "message": { "role": "assistant", "content": "Hello, my name is Litellm Bot. I'm a helpful assistant here to provide information and answer your questions.\n\nTo check the weather for you, I'll need to know your location. Could you please provide me with your city or zip code? That way, I can give you the most accurate and up-to-date weather information.\n\nIf you don't have your location handy, I can also suggest some popular weather websites or apps that you can use to check the weather for your area.\n\nLet me know how I can assist you!", "tool_calls": [], }, "logprobs": None, "finish_reason": "stop", "stop_reason": None, } ], "usage": {"prompt_tokens": 63, "total_tokens": 172, "completion_tokens": 109}, "prompt_logprobs": None, } # Setup mock request mock_request = respx_mock.post(expected_url).mock( return_value=httpx.Response(200, json=mock_response) ) # Make request response = await litellm.acompletion( model="vertex_ai/openai/5464397967697903616", messages=messages, vertex_project="633608382793", vertex_location="us-central1", max_retries=max_retries, ) # Assert request was made correctly assert mock_request.called request_body = json.loads(mock_request.calls[0].request.content) assert request_body == expected_request # Assert response structure assert response.id == "chat-09940d4e99e3488aa52a6f5e2ecf35b1" assert response.created == 1731702782 assert response.model == "vertex_ai/meta-llama/Llama-3.1-8B-Instruct" assert len(response.choices) == 1 assert response.choices[0].message.role == "assistant" assert response.choices[0].message.content.startswith( "Hello, my name is Litellm Bot" ) assert response.choices[0].finish_reason == "stop" assert response.usage.completion_tokens == 109 assert response.usage.prompt_tokens == 63 assert response.usage.total_tokens == 172 def vertex_ai_anthropic_thinking_mock_response(*args, **kwargs): mock_response = MagicMock() mock_response.status_code = 200 mock_response.headers = {"Content-Type": "application/json"} mock_response.json.return_value = { "id": "msg_vrtx_011pL6Np3MKxXL3R8theMRJW", "type": "message", "role": "assistant", "model": "claude-4-sonnet-20250514", "content": [ { "type": "thinking", "thinking": 'This is a very simple and common greeting in programming and computing. "Hello, world!" is often the first program people write when learning a new programming language, where they create a program that outputs this phrase.\n\nI should respond in a friendly way and acknowledge this greeting. I can keep it simple and welcoming.', "signature": "EugBCkYQAhgCIkAqCkezmsp8DG9Jjoc/CD7yXavPXVvP4TAuwjc/ZgHRIgroz5FzAYxic3CnNiW5w2fx/4+1f4ZYVxWJVLmrEA46EgwFsxbpN2jxMxjIzy0aDIAbMy9rW6B5lGVETCIw4r2UW0A7m5Df991SMSMPvHU9VdL8p9S/F2wajLnLVpl5tH89csm4NqnMpxnou61yKlCLldFGIto1Kvit5W1jqn2gx2dGIOyR4YaJ0c8AIFfQa5TIXf+EChVDzhPKLWZ8D/Q3gCGxBx+m/4dLI8HMZA8Ob3iCMI23eBKmh62FCWJGuA==", }, { "type": "text", "text": "Hi there! 👋 \n\nIt's nice to meet you! \"Hello, world!\" is such a classic phrase in computing - it's often the first output from someone's very first program.\n\nHow are you doing today? Is there something specific I can help you with?", }, ], "stop_reason": "end_turn", "stop_sequence": None, "usage": { "input_tokens": 39, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 0, "output_tokens": 134, }, } return mock_response def test_vertex_anthropic_completion(): from litellm import completion from litellm.llms.custom_httpx.http_handler import HTTPHandler client = HTTPHandler() load_vertex_ai_credentials() with patch.object( client, "post", side_effect=vertex_ai_anthropic_thinking_mock_response ): response = completion( model="vertex_ai/claude-sonnet-4-6@default", messages=[{"role": "user", "content": "Hello, world!"}], vertex_ai_location="us-east5", vertex_ai_project="test-project", thinking={"type": "enabled", "budget_tokens": 1024}, client=client, ) print(response) assert response.model == "claude-sonnet-4-6@default" assert response._hidden_params["response_cost"] is not None assert response._hidden_params["response_cost"] > 0 assert response.choices[0].message.reasoning_content is not None assert isinstance(response.choices[0].message.reasoning_content, str) assert response.choices[0].message.thinking_blocks is not None assert isinstance(response.choices[0].message.thinking_blocks, list) assert len(response.choices[0].message.thinking_blocks) > 0 def test_signed_s3_url_with_format(): from litellm import completion from litellm.llms.custom_httpx.http_handler import HTTPHandler client = HTTPHandler() load_vertex_ai_credentials() args = { "model": "vertex_ai/gemini-2.0-flash-001", "messages": [ { "role": "user", "content": [ { "type": "image_url", "image_url": { "url": "https://litellm-logo-aws-marketplace.s3.us-west-2.amazonaws.com/berriai-logo-github.png?response-content-disposition=inline&X-Amz-Content-Sha256=UNSIGNED-PAYLOAD&X-Amz-Security-Token=IQoJb3JpZ2luX2VjENj%2F%2F%2F%2F%2F%2F%2F%2F%2F%2FwEaCXVzLXdlc3QtMiJGMEQCIHlAy6QneghdEo4Dp4rw%2BHhdInKX4MU3T0hZT1qV3AD%2FAiBGY%2FtfxmBJkj%2BK6%2FxAgek6L3tpOcq6su1mBrj87El%2FCirLAwghEAEaDDg4ODYwMjIyMzQyOCIMzds7lsxAFHHCRHmkKqgDgnsJBaEmmwXBWqzyMMe3BUKsCqfvrYupFGxBREP%2BaEz%2ByLSKiTM3xWzaRz6vrP9T4HSJ97B9wQ3dhUBT22XzdOFsaq49wZapwy9hoPNrMyZ77DIa0MlEbg0uudGOaMAw4NbVEqoERQuZmIMMbNHCeoJsZxKCttRZlTDzU%2FeNNy96ltb%2FuIkX5b3OOYdUaKj%2FUjmPz%2FEufY%2Bn%2FFHawunSYXJwL4pYuBF1IKRtPjqamaYscH%2FrzD7fubGUMqk6hvyGEo%2BLqnVyruQEmVFqAnXyWlpHGqeWazEC7xcsC2lhLO%2FKUouyVML%2FxyYtL4CuKp52qtLWWauAFGnyBZnCHtSL58KLaMTSh7inhoFFIKDN2hymrJ4D9%2Bxv%2FMOzefH5X%2B0pcdJUwyxcwgL3myggRmIYq1L6IL4I%2F54BIU%2FMctJcRXQ8NhQNP2PsaCsXYHHVMXRZxps9v8t9Ciorb0PAaLr0DIGVgEqejSjwbzNTctQf59Rj0GhZ0A6A3nFaq3nL4UvO51aPP6aelN6RnLwHh8fF80iPWII7Oj9PWn9bkON%2F7%2B5k42oPFR0KDTD0yaO%2BBjrlAouRvkyHZnCuLuJdEeqc8%2Fwm4W8SbMiYDzIEPPe2wFR2sH4%2FDlnJRqia9Or00d4N%2BOefBkPv%2Bcdt68r%2FwjeWOrulczzLGjJE%2FGw1Lb9dtGtmupGm2XKOW3geJwXkk1qcr7u5zwy6DNamLJbitB026JFKorRnPajhe5axEDv%2BRu6l1f0eailIrCwZ2iytA94Ni8LTha2GbZvX7fFHcmtyNlgJPpMcELdkOEGTCNBldGck5MFHG27xrVrlR%2F7HZIkKYlImNmsOIjuK7acDiangvVdB6GlmVbzNUKtJ7YJhS2ivwvdDIf8XuaFAkhjRNpewDl0GzPvojK%2BDTizZydyJL%2B20pVkSXptyPwrrHEeiOFWwhszW2iTZij4rlRAoZW6NEdfkWsXrGMbxJTZa3E5URejJbg%2B4QgGtjLrgJhRC1pJGP02GX7VMxVWZzomfC2Hn7WaF44wgcuqjE4HGJfpA2ZLBxde52g%3D%3D&X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=ASIA45ZGR4NCKIUOODV3%2F20250305%2Fus-west-2%2Fs3%2Faws4_request&X-Amz-Date=20250305T235823Z&X-Amz-Expires=43200&X-Amz-SignedHeaders=host&X-Amz-Signature=71a900a9467eaf3811553500aaf509a10a9e743a8133cfb6a78dcbcbc6da4a05", "format": "image/jpeg", }, }, {"type": "text", "text": "Describe this image"}, ], } ], } with patch.object(client, "post", new=MagicMock()) as mock_client: try: response = completion(**args, client=client) print(response) except Exception as e: print(e) print(mock_client.call_args.kwargs) mock_client.assert_called() print(mock_client.call_args.kwargs) json_str = json.dumps(mock_client.call_args.kwargs["json"]) assert "image/jpeg" in json_str assert "image/png" not in json_str def test_gemini_fine_tuned_model_request_consistency(): """ Assert the same transformation is applied to Fine tuned gemini 2.0 flash and gemini 2.0 flash - Request 1: Fine tuned: vertex_ai/gemini/ft-uuid - Request 2: vertex_ai/gemini-2.5-flash """ litellm.set_verbose = True load_vertex_ai_credentials() from unittest.mock import MagicMock, patch from litellm.llms.custom_httpx.http_handler import HTTPHandler # Set up the messages messages = [ { "role": "system", "content": "Your name is Litellm Bot, you are a helpful assistant", }, { "role": "user", "content": "Hello, what is your name and can you tell me the weather?", }, ] # Define tools tools = [ { "type": "function", "function": { "name": "get_weather", "description": "Get the current weather in a given location", "parameters": { "type": "object", "properties": { "location": { "type": "string", "description": "The city and state, e.g. San Francisco, CA", } }, "required": ["location"], }, }, } ] client = HTTPHandler(concurrent_limit=1) # First request with patch.object(client, "post", new=MagicMock()) as mock_post_1: try: response_1 = completion( model="vertex_ai/gemini/ft-uuid", messages=messages, tools=tools, tool_choice="auto", client=client, ) except Exception as e: print(e) # Store the request body from the first call first_request_body = mock_post_1.call_args.kwargs["json"] print("first_request_body", first_request_body) # Validate correct `model` is added to the request to Vertex AI print("final URL=", mock_post_1.call_args.kwargs["url"]) # Validate the request url assert ( "publishers/google/models/ft-uuid:generateContent" in mock_post_1.call_args.kwargs["url"] ) # Second request with patch.object(client, "post", new=MagicMock()) as mock_post_2: try: response_2 = completion( model="vertex_ai/gemini-2.5-flash", messages=messages, tools=tools, tool_choice="auto", client=client, ) except Exception as e: print(e) # Store the request body from the second call second_request_body = mock_post_2.call_args.kwargs["json"] print("second_request_body", second_request_body) # Get the diff between the two request bodies # Convert dictionaries to formatted JSON strings import json first_json = json.dumps(first_request_body, indent=2).splitlines() second_json = json.dumps(second_request_body, indent=2).splitlines() # Assert there is no difference between the request bodies assert first_json == second_json, "Request bodies should be identical" def test_vertex_ai_llama_tool_calling(): os.environ["LITELLM_LOCAL_MODEL_COST_MAP"] = "True" litellm.model_cost = litellm.get_model_cost_map(url="") load_vertex_ai_credentials() litellm.turn_on_debug() args = { "model": "vertex_ai/meta/llama-4-maverick-17b-128e-instruct-maas", "messages": [ {"role": "user", "content": "What is the weather in Boston, MA today?"} ], "tools": [ { "type": "function", "function": { "name": "get_weather", "description": "Get current temperature for a given location.", "parameters": { "type": "object", "properties": { "location": { "type": "string", "description": "City and country e.g. Bogotá, Colombia", } }, "required": ["location"], "additionalProperties": False, }, }, } ], "vertex_location": "us-east5", } try: response = completion(**args) except litellm.RateLimitError: pytest.skip("Rate limit error") except litellm.NotFoundError: pytest.skip("Model not found / resource unavailable") print(response) assert response.choices[0].message.tool_calls is not None assert response.choices[0].finish_reason == "tool_calls" assert response._hidden_params["response_cost"] > 0 def test_gemini_nullable_object_tool_schema_httpx(): """ Ensure nullable object tool params preserve nested properties in Vertex schema conversion. """ load_vertex_ai_credentials() litellm.turn_on_debug() tools = [ { "type": "function", "strict": True, "function": { "name": "create_support_ticket", "description": "Create a paid user support ticket", "parameters": { "type": "object", "additionalProperties": False, "required": ["ticket_id", "customer_context"], "properties": { "ticket_id": { "type": "string", "description": "Unique identifier for the support ticket", }, "customer_context": { "type": ["object", "null"], "description": "Context about the paid customer, if available", "additionalProperties": False, "required": ["user_id", "plan"], "properties": { "user_id": { "type": "string", "description": "Internal user identifier", }, "plan": { "type": "string", "description": "Subscription plan name (e.g. pro, enterprise)", }, }, }, }, }, }, } ] response = litellm.completion( model="vertex_ai/gemini-3.5-flash", messages=[{"role": "user", "content": "call the tool"}], tools=tools, tool_choice="required", vertex_location="global", ) print(response) def test_vertex_ai_response_id(): """Test that litellm preserves the response ID from Vertex AI's API for non-streaming responses""" from litellm.llms.custom_httpx.http_handler import HTTPHandler load_vertex_ai_credentials() client = HTTPHandler() mock_response = MagicMock() mock_response.status_code = 200 mock_response.headers = {"Content-Type": "application/json"} mock_response.json.return_value = { "responseId": "vertex_ai_response_123", "candidates": [ { "content": { "role": "model", "parts": [{"text": "Hello! How can I help you today?"}], }, "finishReason": "STOP", "safetyRatings": [ { "category": "HARM_CATEGORY_HATE_SPEECH", "probability": "NEGLIGIBLE", } ], } ], "usageMetadata": { "promptTokenCount": 10, "candidatesTokenCount": 8, "totalTokenCount": 18, }, } with patch.object(client, "post", return_value=mock_response) as mock_post: response = completion( model="vertex_ai/gemini-1.5-pro", messages=[{"role": "user", "content": "Hi!"}], client=client, ) # Verify the response ID is preserved assert response.id == "vertex_ai_response_123" assert response.choices[0].message.content == "Hello! How can I help you today?" def test_vertex_ai_gemini_2_5_pro_streaming(): try: load_vertex_ai_credentials() # litellm.turn_on_debug() response = completion( model="vertex_ai/gemini-2.5-pro", messages=[{"role": "user", "content": "Hi!"}], vertex_location="global", stream=True, ) has_real_content = False for chunk in response: print(chunk) if ( chunk.choices[0].delta.content is not None and len(chunk.choices[0].delta.content) > 0 ): has_real_content = True assert has_real_content except litellm.RateLimitError: pytest.skip("Skipping due to rate limit error") @pytest.mark.asyncio async def test_vertex_ai_deepseek(): """Test that deepseek models use the correct v1 API endpoint instead of v1beta1.""" load_vertex_ai_credentials() litellm.turn_on_debug() from litellm.llms.custom_httpx.http_handler import AsyncHTTPHandler client = AsyncHTTPHandler() # Create a proper mock response mock_response = MagicMock() mock_response.json.return_value = { "choices": [ { "message": { "role": "assistant", "content": "Hello! How can I help you today?", }, "index": 0, "finish_reason": "stop", } ], "usage": {"prompt_tokens": 10, "completion_tokens": 20, "total_tokens": 30}, "model": "deepseek-ai/deepseek-r1-0528-maas", } mock_response.status_code = 200 with patch.object(client, "post", return_value=mock_response) as mock_post: response = await acompletion( model="vertex_ai/deepseek-ai/deepseek-r1-0528-maas", messages=[{"role": "user", "content": "Hi!"}], client=client, ) mock_post.assert_called_once() # Access the URL from kwargs since the call is made with keyword arguments url = mock_post.call_args.kwargs["url"] print(f"mock_post.call_args.kwargs['url']: {url}") assert "v1beta1" not in url assert "v1" in url def test_gemini_grounding_on_streaming(): from litellm import completion load_vertex_ai_credentials() # litellm.turn_on_debug() args = { "model": "vertex_ai/gemini-3-flash-preview", "messages": [ { "role": "user", "content": [ { "type": "text", "text": "What is the weather like on San Francisco today ?", } ], } ], "vertex_location": "global", "stream": True, "tools": [{"googleSearch": {}}], "fallbacks": [], } result = completion(**args) vertex_ai_grounding_metadata_shows_up = False for chunk in result: if hasattr(chunk, "vertex_ai_grounding_metadata"): vertex_ai_grounding_metadata_shows_up = True print(chunk) assert vertex_ai_grounding_metadata_shows_up def test_gemini_google_maps_tool_simple(): """ Test googleMaps tool with just enableWidget parameter. """ load_vertex_ai_credentials() litellm.turn_on_debug() tools = [{"googleMaps": {"enableWidget": True}}] tools_with_location = [ { "googleMaps": { "enableWidget": True, "latitude": 37.7749, "longitude": -122.4194, "languageCode": "en_US", } } ] try: for tools in [tools, tools_with_location]: response = completion( model="vertex_ai/gemini-3-flash-preview", messages=[ { "role": "user", "content": "What restaurants are nearby?", } ], tools=tools, vertex_location="global", ) print(f"Response: {response.model_dump_json(indent=4)}") assert response.choices[0].message.content is not None except (litellm.RateLimitError, litellm.InternalServerError) as e: pytest.skip(f"Transient Vertex-side failure, not a LiteLLM bug: {e}") except Exception as e: pytest.fail(f"Error occurred: {e}")