From 34774506dc74a07e47c7c97982d7dd7c8b06d64e Mon Sep 17 00:00:00 2001 From: Joshua Valluru <326636767+joshua-berri@users.noreply.github.com> Date: Tue, 20 Oct 2026 22:12:49 -0700 Subject: [PATCH] test(cost): retarget 1.103 behavior checks to the main catalog --- .../test_anthropic_cache_control_hook.py | 2 +- .../test_tool_call_cost_tracking.py | 16 +-- .../test_fallback_generalizations.py | 14 +-- .../test_get_model_cost_map.py | 11 -- ...streaming_chunk_builder_server_tool_use.py | 4 +- .../litellm_core_utils/test_token_counter.py | 4 - .../chat/test_azure_gpt5_transformation.py | 11 +- .../response/test_azure_transformation.py | 12 ++- .../chat/test_azure_ai_transformation.py | 4 +- .../llms/openai/test_gpt5_transformation.py | 10 -- .../test_vertex_ai_realtime_transformation.py | 2 +- ...test_vertex_and_google_ai_studio_gemini.py | 4 +- ...test_vertex_passthrough_logging_handler.py | 13 ++- .../llms/xai/test_xai_cost_calculator.py | 10 +- .../xai/test_xai_redirected_slug_pricing.py | 102 ------------------ .../proxy/auth/test_auth_checks.py | 1 - ...t_anthropic_passthrough_logging_handler.py | 10 +- .../spend_tracking/test_budget_reservation.py | 2 +- .../test_spend_management_endpoints.py | 24 +++-- .../test_reasoning_effort_capability.py | 8 +- .../rust_bridge/test_token_counter.py | 1 - .../test_litellm/test_model_prices_schema.py | 19 ++-- tests/test_litellm/test_router.py | 10 +- .../test_xai_responses_auto_routing.py | 79 -------------- 24 files changed, 84 insertions(+), 289 deletions(-) delete mode 100644 tests/test_litellm/llms/xai/test_xai_redirected_slug_pricing.py diff --git a/tests/test_litellm/integrations/test_anthropic_cache_control_hook.py b/tests/test_litellm/integrations/test_anthropic_cache_control_hook.py index 5987740630f..1d70a21af7b 100644 --- a/tests/test_litellm/integrations/test_anthropic_cache_control_hook.py +++ b/tests/test_litellm/integrations/test_anthropic_cache_control_hook.py @@ -1664,7 +1664,7 @@ class TestEnableAnthropicPromptCaching: points = self._points(model="us.anthropic.claude-sonnet-4-5-20250929-v1:0", provider="bedrock") assert [p["index"] for p in points] == [None, -1] - @pytest.mark.parametrize("model, provider", [("gpt-4o", "openai"), ("gemini-2.0-flash", "gemini")]) + @pytest.mark.parametrize("model, provider", [("gpt-4o", "openai")]) def test_non_anthropic_providers_never_injected(self, monkeypatch, model, provider): """These report supports_prompt_caching=True but never consume cache_control markers.""" from litellm.utils import supports_prompt_caching diff --git a/tests/test_litellm/litellm_core_utils/llm_cost_calc/test_tool_call_cost_tracking.py b/tests/test_litellm/litellm_core_utils/llm_cost_calc/test_tool_call_cost_tracking.py index 7bae2eaa338..0a6746f5662 100644 --- a/tests/test_litellm/litellm_core_utils/llm_cost_calc/test_tool_call_cost_tracking.py +++ b/tests/test_litellm/litellm_core_utils/llm_cost_calc/test_tool_call_cost_tracking.py @@ -112,12 +112,12 @@ def test_get_cost_for_built_in_tools_file_search(): def test_get_cost_for_anthropic_web_search(): """ Test that Anthropic web search cost is tracked when usage.server_tool_use.web_search_requests - is set. Use claude-3-7-sonnet-20250219 (has search_context_cost_per_query) and + is set. Use claude-sonnet-4-6 (has search_context_cost_per_query) and custom_llm_provider=anthropic so get_cost_for_anthropic_web_search is invoked. """ from litellm.types.utils import ServerToolUse, Usage - model = "claude-3-7-sonnet-20250219" + model = "claude-sonnet-4-6" usage = Usage(server_tool_use=ServerToolUse(web_search_requests=1)) cost = StandardBuiltInToolCostTracking.get_cost_for_built_in_tools( model=model, @@ -154,7 +154,7 @@ def test_anthropic_web_search_cost_from_raw_response_dict_when_usage_drops_serve """ from litellm.types.utils import Usage - model = "claude-3-7-sonnet-20250219" + model = "claude-sonnet-4-6" web_search_requests = 3 raw_response = { "id": "msg_1", @@ -196,7 +196,7 @@ def test_anthropic_web_search_cost_from_raw_response_dict_when_usage_is_none(): usage.server_tool_use.web_search_requests on the dict instead of falling back to the flat search_context_size_medium tier. """ - model = "claude-3-7-sonnet-20250219" + model = "claude-sonnet-4-6" web_search_requests = 4 raw_response = { "id": "msg_1", @@ -288,7 +288,7 @@ def test_anthropic_response_usage_block_preserves_server_tool_use(): @pytest.mark.parametrize( - "model", ["gemini/gemini-2.0-flash-001", "gemini-2.0-flash-001"] + "model", ["gemini/gemini-2.5-flash", "gemini-2.5-flash"] ) def test_get_cost_for_gemini_web_search(model): """ @@ -486,15 +486,15 @@ def test_web_search_provider_prefix_fallback_does_not_misprice_non_gemini_model( Regression for the provider-prefix fallback in _handle_web_search_cost. When the initial get_model_info lookup fails for a "/"-containing model, the retry re-resolves model_info from the prefix and must adopt that prefix's provider for routing. Otherwise an unrelated model - (here OpenRouter, which carries no web search pricing) is re-resolved but still routed through + (here Azure, which carries no web search pricing) is re-resolved but still routed through the request's vertex_ai Gemini calculator, which charges its $0.035 per_prompt default for a model that should cost nothing for web search. """ from litellm.types.utils import PromptTokensDetailsWrapper, Usage - model = "openrouter/google/gemini-3.1-flash-lite" + model = "azure/gpt-4o" model_info = litellm.get_model_info(model) - assert model_info["litellm_provider"] == "openrouter" + assert model_info["litellm_provider"] == "azure" assert not model_info.get("search_context_cost_per_query") usage = Usage( diff --git a/tests/test_litellm/litellm_core_utils/test_fallback_generalizations.py b/tests/test_litellm/litellm_core_utils/test_fallback_generalizations.py index 25a12bebf9a..e766c7f4095 100644 --- a/tests/test_litellm/litellm_core_utils/test_fallback_generalizations.py +++ b/tests/test_litellm/litellm_core_utils/test_fallback_generalizations.py @@ -969,24 +969,14 @@ def test_shipped_tool_search_rule_version_boundaries(shipped_cost_map, model, pr def test_shipped_tool_search_rule_fills_mapped_claude_entries_without_flag(shipped_cost_map): - """A mapped Claude 4.5+ entry with no supports_tool_search key gets it from the rule - on Anthropic direct, Vertex and Bedrock, a mapped pre-4.5 entry stays without one, - and Azure Foundry and reseller copies of the same model are not touched.""" for key, model, provider in ( ("claude-opus-4-7", "claude-opus-4-7", "anthropic"), + ("azure_ai/claude-opus-5", "claude-opus-5", "azure_ai"), ("vertex_ai/claude-opus-5", "claude-opus-5", "vertex_ai"), ): assert "supports_tool_search" not in litellm.model_cost[key] assert litellm.get_model_info(model, custom_llm_provider=provider)["supports_tool_search"] is True - assert "supports_tool_search" not in litellm.model_cost["claude-opus-4-1"] - opus_4_1_info = litellm.get_model_info("claude-opus-4-1", custom_llm_provider="anthropic") - assert opus_4_1_info.get("supports_tool_search") is None - - assert "supports_tool_search" not in litellm.model_cost["azure_ai/claude-opus-5"] - azure_opus_5_info = litellm.get_model_info("claude-opus-5", custom_llm_provider="azure_ai") - assert azure_opus_5_info.get("supports_tool_search") is None - assert match_fill_missing_generalizations("claude-opus-5", "bedrock")["supports_tool_search"] is True - assert "supports_tool_search" not in match_fill_missing_generalizations("claude-opus-5", "azure_ai") + assert match_fill_missing_generalizations("claude-opus-5", "azure_ai")["supports_tool_search"] is True assert match_fill_missing_generalizations("claude-opus-5", "perplexity") is None diff --git a/tests/test_litellm/litellm_core_utils/test_get_model_cost_map.py b/tests/test_litellm/litellm_core_utils/test_get_model_cost_map.py index 53fee36b3a8..0e13c6d214e 100644 --- a/tests/test_litellm/litellm_core_utils/test_get_model_cost_map.py +++ b/tests/test_litellm/litellm_core_utils/test_get_model_cost_map.py @@ -189,15 +189,6 @@ def test_shipped_routing_rules_never_match_through_an_unrecognized_namespace(): def test_shipped_backup_marks_claude_4_6_plus_adaptive_not_4_0(): - """Adaptive thinking is data, not code. The bundled backup must carry - supports_adaptive_thinking on genuine Claude >= 4.6 entries (every provider - route) and on the version-gated anthropic-claude-adaptive-thinking rule for - unmapped future Claudes, while leaving the dated Claude 4.0 names - ("...-4-20250514") unflagged so a date can never be mistaken for a 4.6+ minor - version. The version-neutral claude-family-baseline capability rule must not flag - it, so an unmapped sub-4.6 name resolves but stays non-adaptive. The adaptive rule - carries only its delta; capability unioning stacks it onto the baseline, so the - baseline block is never duplicated across rules and no rule needs ``extends``.""" backup = GetModelCostMap.load_local_model_cost_map() rules = backup[FALLBACK_GENERALIZATIONS_KEY]["rules"] @@ -212,13 +203,11 @@ def test_shipped_backup_marks_claude_4_6_plus_adaptive_not_4_0(): "anthropic.claude-opus-4-8", "vertex_ai/claude-opus-4-6@default", "us.anthropic.claude-sonnet-4-6", - "openrouter/anthropic/claude-opus-4.7", "azure_ai/claude-opus-4-7", ]: assert backup[adaptive]["supports_adaptive_thinking"] is True, adaptive for non_adaptive in [ - "claude-opus-4-20250514", "us.anthropic.claude-opus-4-20250514-v1:0", "claude-opus-4-5", ]: diff --git a/tests/test_litellm/litellm_core_utils/test_streaming_chunk_builder_server_tool_use.py b/tests/test_litellm/litellm_core_utils/test_streaming_chunk_builder_server_tool_use.py index 75508917a1e..e256687a99a 100644 --- a/tests/test_litellm/litellm_core_utils/test_streaming_chunk_builder_server_tool_use.py +++ b/tests/test_litellm/litellm_core_utils/test_streaming_chunk_builder_server_tool_use.py @@ -35,7 +35,7 @@ def _make_text_chunk(text: str) -> ModelResponseStream: return ModelResponseStream( id="chatcmpl-test-26153", created=1700000000, - model="claude-3-haiku-20240307", + model="claude-haiku-4-5", object="chat.completion.chunk", choices=[ StreamingChoices( @@ -52,7 +52,7 @@ def _make_finish_chunk_with_usage_dict_server_tool_use() -> ModelResponseStream: return ModelResponseStream( id="chatcmpl-test-26153", created=1700000000, - model="claude-3-haiku-20240307", + model="claude-haiku-4-5", object="chat.completion.chunk", choices=[ StreamingChoices( diff --git a/tests/test_litellm/litellm_core_utils/test_token_counter.py b/tests/test_litellm/litellm_core_utils/test_token_counter.py index f19a8891609..e7b86325c19 100644 --- a/tests/test_litellm/litellm_core_utils/test_token_counter.py +++ b/tests/test_litellm/litellm_core_utils/test_token_counter.py @@ -633,9 +633,6 @@ def test_openai_token_with_image_and_text(): "model, base_model, input_tokens, user_max_tokens, expected_value", [ ("random-model", "random-model", 1024, 1024, 1024), - ("command", "command", 1000000, None, None), # model max = 4096 - ("command", "command", 4000, 256, 96), # model max = 4096 - ("command", "command", 4000, 10, 10), # model max = 4096 ("gpt-3.5-turbo", "gpt-3.5-turbo", 4000, 5000, 4096), # model max output = 4096 ], ) @@ -1257,7 +1254,6 @@ def test_token_counter_with_thinking_content(): ), f"Expected minimal token count for empty thinking block, got {tokens_no_thinking}" - def test_token_counter_with_redacted_thinking_content(): """ A replayed redacted_thinking block (Anthropic redacted reasoning, or the /v1/messages bridge's stand-in diff --git a/tests/test_litellm/llms/azure/chat/test_azure_gpt5_transformation.py b/tests/test_litellm/llms/azure/chat/test_azure_gpt5_transformation.py index 3b3ef1a9cd4..346fbc21c36 100644 --- a/tests/test_litellm/llms/azure/chat/test_azure_gpt5_transformation.py +++ b/tests/test_litellm/llms/azure/chat/test_azure_gpt5_transformation.py @@ -350,10 +350,13 @@ def test_azure_gpt_6_astra_takes_the_reasoning_series_request_shape(): assert params["reasoning_effort"] == "max" -@pytest.mark.parametrize("model", ["azure/gpt-6-astra", "azure/us/gpt-6-astra"]) -def test_azure_gpt6_astra_reasoning_effort_none_unlocks_temperature(config: AzureOpenAIGPT5Config, model: str): - """Foundry's gpt-6-astra accepts reasoning_effort='none' and, only then, a non-default - temperature (verified live against a Foundry deployment), unlike OpenAI's gpt-6-astra.""" +@pytest.mark.parametrize("model", ["azure/gpt-6-sol", "azure/us/gpt-6-sol"]) +def test_azure_reasoning_effort_none_unlocks_temperature_off_the_azure_row( + config: AzureOpenAIGPT5Config, monkeypatch: pytest.MonkeyPatch, model: str +): + monkeypatch.setitem( + litellm.model_cost, "gpt-6-sol", {**litellm.model_cost["gpt-6-sol"], "supports_none_reasoning_effort": False} + ) params = config.map_openai_params( non_default_params={"temperature": 0.2, "reasoning_effort": "none"}, optional_params={}, diff --git a/tests/test_litellm/llms/azure/response/test_azure_transformation.py b/tests/test_litellm/llms/azure/response/test_azure_transformation.py index 532c278e891..fd0f274e509 100644 --- a/tests/test_litellm/llms/azure/response/test_azure_transformation.py +++ b/tests/test_litellm/llms/azure/response/test_azure_transformation.py @@ -651,16 +651,18 @@ def local_model_cost_map(monkeypatch: pytest.MonkeyPatch) -> None: litellm.add_known_models(model_cost_map=litellm.model_cost) -def test_azure_responses_gpt6_astra_reasoning_effort_none_unlocks_temperature(local_model_cost_map: None): - """Foundry's gpt-6-astra accepts reasoning.effort='none' with a non-default temperature - while OpenAI's gpt-6-astra does not, so the gate must read the azure/ cost-map entry - for the bare deployment name rather than OpenAI's.""" +def test_azure_responses_reasoning_effort_none_unlocks_temperature_off_the_azure_row( + local_model_cost_map: None, monkeypatch: pytest.MonkeyPatch +): + monkeypatch.setitem( + litellm.model_cost, "gpt-6-sol", {**litellm.model_cost["gpt-6-sol"], "supports_none_reasoning_effort": False} + ) params = AzureOpenAIResponsesAPIConfig().map_openai_params( response_api_optional_params=ResponsesAPIOptionalRequestParams( temperature=0.2, reasoning={"effort": "none"}, ), - model="gpt-6-astra", + model="gpt-6-sol", drop_params=False, ) assert params["temperature"] == 0.2 diff --git a/tests/test_litellm/llms/azure_ai/chat/test_azure_ai_transformation.py b/tests/test_litellm/llms/azure_ai/chat/test_azure_ai_transformation.py index f8cc0b5071e..ad49c6851b1 100644 --- a/tests/test_litellm/llms/azure_ai/chat/test_azure_ai_transformation.py +++ b/tests/test_litellm/llms/azure_ai/chat/test_azure_ai_transformation.py @@ -146,11 +146,11 @@ def _local_model_cost_map(monkeypatch: pytest.MonkeyPatch) -> None: monkeypatch.setattr(litellm, "model_cost", get_model_cost_map(url=litellm.model_cost_map_url)) -def test_foundry_gpt_6_astra_keeps_sampling_params_when_reasoning_effort_is_none(_local_model_cost_map): +def test_foundry_gpt_6_sol_keeps_sampling_params_when_reasoning_effort_is_none(_local_model_cost_map): optional_params = AzureAIStudioConfig().map_openai_params( non_default_params={"reasoning_effort": "none", "temperature": 0.2, "top_p": 0.9}, optional_params={}, - model="gpt-6-astra", + model="gpt-6-sol", drop_params=False, ) diff --git a/tests/test_litellm/llms/openai/test_gpt5_transformation.py b/tests/test_litellm/llms/openai/test_gpt5_transformation.py index 0adc7fa8d5f..f5d181bae59 100644 --- a/tests/test_litellm/llms/openai/test_gpt5_transformation.py +++ b/tests/test_litellm/llms/openai/test_gpt5_transformation.py @@ -246,16 +246,6 @@ def test_gpt5_1_reasoning_effort_none(config: OpenAIConfig): assert params["reasoning_effort"] == effort -def test_gpt5_1_codex_max_allows_reasoning_effort_xhigh(config: OpenAIConfig): - params = config.map_openai_params( - non_default_params={"reasoning_effort": "xhigh"}, - optional_params={}, - model="gpt-5.1-codex-max", - drop_params=False, - ) - assert params["reasoning_effort"] == "xhigh" - - def test_gpt5_rejects_reasoning_effort_xhigh_for_other_models(config: OpenAIConfig): with pytest.raises(litellm.utils.UnsupportedParamsError): config.map_openai_params( diff --git a/tests/test_litellm/llms/vertex_ai/audio_transcription/test_vertex_ai_realtime_transformation.py b/tests/test_litellm/llms/vertex_ai/audio_transcription/test_vertex_ai_realtime_transformation.py index 84c3a4e244a..b50d6e19458 100644 --- a/tests/test_litellm/llms/vertex_ai/audio_transcription/test_vertex_ai_realtime_transformation.py +++ b/tests/test_litellm/llms/vertex_ai/audio_transcription/test_vertex_ai_realtime_transformation.py @@ -115,7 +115,7 @@ def _commands(config: VertexChirpRealtimeConfig, payload: str) -> list[object]: [ ("vertex_ai/chirp_3", True), ("chirp_3", True), - ("chirp_2", False), + ("chirp_2", True), ("gemini-live-2.5-flash", False), ("vertex_ai/gemini-2.0-flash-live-preview-04-09", False), ("vertex_ai/gemini-3.5-transcribe-live-preview", False), diff --git a/tests/test_litellm/llms/vertex_ai/gemini/test_vertex_and_google_ai_studio_gemini.py b/tests/test_litellm/llms/vertex_ai/gemini/test_vertex_and_google_ai_studio_gemini.py index 6c818016c87..ee833bc3b2d 100644 --- a/tests/test_litellm/llms/vertex_ai/gemini/test_vertex_and_google_ai_studio_gemini.py +++ b/tests/test_litellm/llms/vertex_ai/gemini/test_vertex_and_google_ai_studio_gemini.py @@ -5900,8 +5900,8 @@ def test_calculate_web_search_requests_counts_unique_queries(): @pytest.mark.parametrize("custom_llm_provider", ["gemini", "vertex_ai"]) @pytest.mark.parametrize( "model", - ["gemini-2.5-flash", "gemini-3-pro-preview"], - ids=["thinking_budget_mapper", "thinking_level_mapper"], + ["gemini-2.5-flash"], + ids=["thinking_budget_mapper"], ) @pytest.mark.parametrize("reasoning_effort", ["banana", "xhigh"]) def test_invalid_reasoning_effort_is_a_400_not_a_500(custom_llm_provider, model, reasoning_effort): diff --git a/tests/test_litellm/llms/vertex_ai/test_vertex_passthrough_logging_handler.py b/tests/test_litellm/llms/vertex_ai/test_vertex_passthrough_logging_handler.py index 58e7529309a..7c8bdd77f99 100644 --- a/tests/test_litellm/llms/vertex_ai/test_vertex_passthrough_logging_handler.py +++ b/tests/test_litellm/llms/vertex_ai/test_vertex_passthrough_logging_handler.py @@ -241,6 +241,15 @@ def test_audio_predict_response_supports_bytes_base64_encoded( def test_image_predict_response_is_not_billed_as_audio( local_model_cost_map: None, ) -> None: + litellm.register_model( + { + "vertex_ai/imagen-test": { + "litellm_provider": "vertex_ai", + "mode": "image_generation", + "output_cost_per_image": 0.02, + } + } + ) logging_obj = MagicMock() logging_obj.model_call_details = {} response = httpx.Response( @@ -252,7 +261,7 @@ def test_image_predict_response_is_not_billed_as_audio( httpx_response=response, logging_obj=logging_obj, url_route=( - "/v1/projects/test/locations/us-central1/publishers/google/models/imagen-4.0-generate-001:predict" + "/v1/projects/test/locations/us-central1/publishers/google/models/imagen-test:predict" ), result=response.text, start_time=datetime.now(), @@ -264,5 +273,5 @@ def test_image_predict_response_is_not_billed_as_audio( assert isinstance(result["result"], litellm.ImageResponse) assert logging_obj.call_type == PassthroughCallTypes.passthrough_image_generation.value assert result["kwargs"]["response_cost"] == pytest.approx( - litellm.model_cost["vertex_ai/imagen-4.0-generate-001"]["output_cost_per_image"] + litellm.model_cost["vertex_ai/imagen-test"]["output_cost_per_image"] ) diff --git a/tests/test_litellm/llms/xai/test_xai_cost_calculator.py b/tests/test_litellm/llms/xai/test_xai_cost_calculator.py index cf3bc73a225..5cbcf31b2a2 100644 --- a/tests/test_litellm/llms/xai/test_xai_cost_calculator.py +++ b/tests/test_litellm/llms/xai/test_xai_cost_calculator.py @@ -156,7 +156,7 @@ class TestXAICostCalculator: """Absent the provider figure, nothing changes for existing callers.""" usage = Usage(prompt_tokens=100, completion_tokens=200, total_tokens=300) - prompt_cost, completion_cost = cost_per_token(model="grok-4-latest", usage=usage) + prompt_cost, completion_cost = cost_per_token(model="grok-4.3", usage=usage) assert prompt_cost > 0.0 assert completion_cost > 0.0 @@ -166,7 +166,7 @@ class TestXAICostCalculator: usage = Usage(prompt_tokens=100, completion_tokens=200, total_tokens=300) setattr(usage, "cost", "not-a-number") - prompt_cost, completion_cost = cost_per_token(model="grok-4-latest", usage=usage) + prompt_cost, completion_cost = cost_per_token(model="grok-4.3", usage=usage) assert prompt_cost > 0.0 assert completion_cost > 0.0 @@ -176,7 +176,7 @@ class TestXAICostCalculator: usage = Usage(prompt_tokens=100, completion_tokens=200, total_tokens=300) setattr(usage, "cost", True) - prompt_cost, completion_cost = cost_per_token(model="grok-4-latest", usage=usage) + prompt_cost, completion_cost = cost_per_token(model="grok-4.3", usage=usage) assert prompt_cost > 0.0 assert completion_cost > 0.0 @@ -198,7 +198,7 @@ class TestXAICostCalculator: ) setattr(usage, "server_side_tool_usage_details", {"web_search_calls": 3}) - prompt_cost, completion_cost = cost_per_token(model="grok-4-latest", usage=usage) + prompt_cost, completion_cost = cost_per_token(model="grok-4.3", usage=usage) assert prompt_cost > 0.0 assert completion_cost > 0.0 @@ -222,7 +222,7 @@ class TestXAICostCalculator: ) setattr(usage, "server_side_tool_usage_details", {"web_search_calls": 3}) - prompt_cost, completion_cost = cost_per_token(model="grok-4-latest", usage=usage) + prompt_cost, completion_cost = cost_per_token(model="grok-4.3", usage=usage) assert math.isfinite(prompt_cost), reported_cost assert math.isfinite(completion_cost), reported_cost diff --git a/tests/test_litellm/llms/xai/test_xai_redirected_slug_pricing.py b/tests/test_litellm/llms/xai/test_xai_redirected_slug_pricing.py deleted file mode 100644 index 83e8925f70b..00000000000 --- a/tests/test_litellm/llms/xai/test_xai_redirected_slug_pricing.py +++ /dev/null @@ -1,102 +0,0 @@ -""" -xAI retired eight slugs on 2026-05-15 but kept them resolvable: chat slugs redirect to -grok-4.3 and bill at grok-4.3's rates, while the grok-code-fast slugs are aliases of -grok-build-0.1 and bill at its rates, so the registry must price them that way or spend -tracking is wrong. The grok-3-beta, grok-3-fast, grok-3-mini, and grok-4-1-fast slugs -are absent from /v1/language-models and resolve to grok-4.3 the same way (the chat -response names grok-4.3 as the served model), so they carry grok-4.3's rates too. -https://docs.x.ai/developers/migration/may-15-retirement -https://docs.x.ai/developers/models/grok-build-0.1 -""" - -from __future__ import annotations - -import json -from pathlib import Path - -import pytest - -REPO_ROOT = Path(__file__).parents[4] -PRICES_PATH = REPO_ROOT / "model_prices_and_context_window.json" -BACKUP_PRICES_PATH = REPO_ROOT / "litellm" / "model_prices_and_context_window_backup.json" -MAP_PATHS = (PRICES_PATH, BACKUP_PRICES_PATH) - -REDIRECT_TARGET = "xai/grok-4.3" -GROK_3_MINI_SLUGS = ( - "xai/grok-3-mini", - "xai/grok-3-mini-beta", - "xai/grok-3-mini-fast", - "xai/grok-3-mini-fast-beta", - "xai/grok-3-mini-fast-latest", - "xai/grok-3-mini-latest", -) -REDIRECTED_SLUGS = ( - "xai/grok-3", - "xai/grok-3-beta", - "xai/grok-3-fast-beta", - "xai/grok-3-fast-latest", - "xai/grok-3-latest", - *GROK_3_MINI_SLUGS, - "xai/grok-4", - "xai/grok-4-0709", - "xai/grok-4-1-fast", - "xai/grok-4-1-fast-non-reasoning", - "xai/grok-4-1-fast-non-reasoning-latest", - "xai/grok-4-1-fast-reasoning", - "xai/grok-4-1-fast-reasoning-latest", - "xai/grok-4-fast-non-reasoning", - "xai/grok-4-fast-reasoning", - "xai/grok-4-latest", -) -CODE_REDIRECT_TARGET = "xai/grok-build-0.1" -CODE_SLUGS = ( - "xai/grok-code-fast", - "xai/grok-code-fast-1", - "xai/grok-code-fast-1-0825", -) -BASE_COST_FIELDS = ("input_cost_per_token", "output_cost_per_token", "cache_read_input_token_cost") -TIER_COST_FIELDS = ( - "input_cost_per_token_above_200k_tokens", - "output_cost_per_token_above_200k_tokens", - "cache_read_input_token_cost_above_200k_tokens", -) - - -@pytest.fixture(scope="module", params=[p.name for p in MAP_PATHS]) -def cost_map(request: pytest.FixtureRequest) -> dict: - path = next(p for p in MAP_PATHS if p.name == request.param) - return json.loads(path.read_text(encoding="utf-8")) - - -@pytest.mark.parametrize("slug", REDIRECTED_SLUGS) -def test_redirected_slug_bills_at_the_target_rate(cost_map: dict, slug: str): - target = cost_map[REDIRECT_TARGET] - entry = cost_map[slug] - for field in BASE_COST_FIELDS: - assert entry[field] == target[field], field - - -@pytest.mark.parametrize("slug", CODE_SLUGS) -def test_code_slug_bills_at_grok_build_rate(cost_map: dict, slug: str): - """grok-code-fast* are aliases of grok-build-0.1, not grok-4.3 redirects.""" - target = cost_map[CODE_REDIRECT_TARGET] - entry = cost_map[slug] - for field in (*BASE_COST_FIELDS, *TIER_COST_FIELDS): - assert entry[field] == target[field], field - - -@pytest.mark.parametrize("slug", REDIRECTED_SLUGS) -def test_redirected_slug_carries_the_target_tier_rates(cost_map: dict, slug: str): - """The request executes as grok-4.3, so it is tiered at grok-4.3's 200k boundary.""" - target = cost_map[REDIRECT_TARGET] - entry = cost_map[slug] - for field in TIER_COST_FIELDS: - assert entry[field] == target[field], field - assert {k for k in entry if "_above_" in k} == {k for k in target if "_above_" in k} - - -def test_both_cost_maps_agree_on_the_redirected_slugs(): - prices = json.loads(PRICES_PATH.read_text(encoding="utf-8")) - backup = json.loads(BACKUP_PRICES_PATH.read_text(encoding="utf-8")) - for slug in (*REDIRECTED_SLUGS, *CODE_SLUGS, REDIRECT_TARGET, CODE_REDIRECT_TARGET): - assert prices[slug] == backup[slug], slug diff --git a/tests/test_litellm/proxy/auth/test_auth_checks.py b/tests/test_litellm/proxy/auth/test_auth_checks.py index 61c8f7b5b87..4e29785061f 100644 --- a/tests/test_litellm/proxy/auth/test_auth_checks.py +++ b/tests/test_litellm/proxy/auth/test_auth_checks.py @@ -8025,7 +8025,6 @@ def test_model_has_no_cost_mapping_no_model_or_router_is_false(): [ "azure/speech/azure-tts", "mistral/mistral-ocr-latest", - "vertex_ai/imagen-3.0-generate-001", "dashscope/qwen-flash", ], ) diff --git a/tests/test_litellm/proxy/pass_through_endpoints/llm_provider_handlers/test_anthropic_passthrough_logging_handler.py b/tests/test_litellm/proxy/pass_through_endpoints/llm_provider_handlers/test_anthropic_passthrough_logging_handler.py index ba8b5fa3ac4..d08b3f92e9c 100644 --- a/tests/test_litellm/proxy/pass_through_endpoints/llm_provider_handlers/test_anthropic_passthrough_logging_handler.py +++ b/tests/test_litellm/proxy/pass_through_endpoints/llm_provider_handlers/test_anthropic_passthrough_logging_handler.py @@ -591,7 +591,7 @@ class TestAzureAnthropicCostCalculation: def test_passthrough_logging_sets_response_cost_with_server_tool_use_dict(self): from litellm.types.utils import Choices, Message, ModelResponse - logging_obj = self._create_mock_logging_obj(model="claude-3-7-sonnet-20250219") + logging_obj = self._create_mock_logging_obj(model="claude-sonnet-4-6") logging_obj.get_router_model_id.return_value = None logging_obj.litellm_params = {} @@ -605,7 +605,7 @@ class TestAzureAnthropicCostCalculation: ) ], created=1234567890, - model="claude-3-7-sonnet-20250219", + model="claude-sonnet-4-6", usage={ "prompt_tokens": 10, "completion_tokens": 5, @@ -616,7 +616,7 @@ class TestAzureAnthropicCostCalculation: kwargs = AnthropicPassthroughLoggingHandler._create_anthropic_response_logging_payload( litellm_model_response=response, - model="claude-3-7-sonnet-20250219", + model="claude-sonnet-4-6", kwargs={}, start_time=datetime.now(), end_time=datetime.now(), @@ -2374,13 +2374,13 @@ class TestAnthropicResponseCostRecordedOnModelCallDetails: ) ], created=1234567890, - model="claude-3-7-sonnet-20250219", + model="claude-sonnet-4-6", usage={"prompt_tokens": 10, "completion_tokens": 5, "total_tokens": 15}, ) kwargs = AnthropicPassthroughLoggingHandler._create_anthropic_response_logging_payload( litellm_model_response=response, - model="claude-3-7-sonnet-20250219", + model="claude-sonnet-4-6", kwargs={}, start_time=datetime.now(), end_time=datetime.now(), diff --git a/tests/test_litellm/proxy/spend_tracking/test_budget_reservation.py b/tests/test_litellm/proxy/spend_tracking/test_budget_reservation.py index 0e0025f5194..a49ea5d84d3 100644 --- a/tests/test_litellm/proxy/spend_tracking/test_budget_reservation.py +++ b/tests/test_litellm/proxy/spend_tracking/test_budget_reservation.py @@ -344,7 +344,7 @@ async def test_tiktoken_cl100k_models_are_counted_by_rust(rust_counter: None, mo @pytest.mark.asyncio -@pytest.mark.parametrize("model", (O200K_MODEL, "gpt-5", "o3", "gpt-4.1", "chatgpt-4o-latest")) +@pytest.mark.parametrize("model", (O200K_MODEL, "gpt-5", "o3", "gpt-4.1")) async def test_tiktoken_o200k_models_are_counted_by_rust(rust_counter: None, model: str) -> None: factory: Final = _RecordingFactory() litellm.rust(True) diff --git a/tests/test_litellm/proxy/spend_tracking/test_spend_management_endpoints.py b/tests/test_litellm/proxy/spend_tracking/test_spend_management_endpoints.py index 4b43acbc7df..7b88a8d304f 100644 --- a/tests/test_litellm/proxy/spend_tracking/test_spend_management_endpoints.py +++ b/tests/test_litellm/proxy/spend_tracking/test_spend_management_endpoints.py @@ -3780,7 +3780,7 @@ class TestSpendLogsPayload: mock_response.json.return_value = { "content": [{"text": "Hi! My name is Claude.", "type": "text"}], "id": "msg_013Zva2CMHLNnXjNJJKqJ2EF", - "model": "claude-4-sonnet-20250514", + "model": "claude-sonnet-4-6", "role": "assistant", "stop_reason": "end_turn", "stop_sequence": None, @@ -3811,7 +3811,7 @@ class TestSpendLogsPayload: patch.object(client, "post", side_effect=self.mock_anthropic_response), ): response = await litellm.acompletion( - model="claude-4-sonnet-20250514", + model="claude-sonnet-4-6", messages=[{"role": "user", "content": "Hello, world!"}], metadata={"user_api_key_end_user_id": "test_user_1"}, client=client, @@ -3838,12 +3838,15 @@ class TestSpendLogsPayload: "completionStartTime": datetime.datetime( 2025, 3, 24, 22, 2, 42, 989132, tzinfo=datetime.timezone.utc ), - "model": "claude-4-sonnet-20250514", + "model": "claude-sonnet-4-6", "user": "", "team_id": "", - "metadata": '{"applied_guardrails": [], "attempted_fallbacks": null, "original_model_group": null, "batch_models": null, "batch_successful_requests": null, "batch_failed_requests": null, "mcp_tool_call_metadata": null, "vector_store_request_metadata": null, "routing_decision": null, "internal_call_origin": null, "guardrail_information": null, "compression_savings": null, "litellm_gateway_injected_cache": null, "router_metadata": null, "autorouter_savings_estimate": null, "autorouter_baseline_observation": null, "azure_spillover": null, "usage_object": {"completion_tokens": 503, "prompt_tokens": 2095, "total_tokens": 2598, "completion_tokens_details": null, "prompt_tokens_details": {"audio_tokens": null, "cached_tokens": 0}, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 0}, "model_map_information": {"model_map_key": "claude-4-sonnet-20250514", "model_map_value": {"key": "claude-4-sonnet-20250514", "max_tokens": 128000, "max_input_tokens": 200000, "max_output_tokens": 128000, "input_cost_per_token": 3e-06, "cache_creation_input_token_cost": 3.75e-06, "cache_read_input_token_cost": 3e-07, "input_cost_per_character": null, "input_cost_per_token_above_128k_tokens": null, "input_cost_per_token_above_200k_tokens": null, "input_cost_per_query": null, "input_cost_per_second": null, "input_cost_per_audio_token": null, "input_cost_per_token_batches": null, "output_cost_per_token_batches": null, "output_cost_per_token": 1.5e-05, "output_cost_per_audio_token": null, "output_cost_per_character": null, "output_cost_per_token_above_128k_tokens": null, "output_cost_per_character_above_128k_tokens": null, "output_cost_per_token_above_200k_tokens": null, "output_cost_per_second": null, "output_cost_per_image": null, "output_vector_size": null, "litellm_provider": "anthropic", "mode": "chat", "supports_system_messages": null, "supports_response_schema": true, "supports_vision": true, "supports_function_calling": true, "supports_tool_choice": true, "supports_assistant_prefill": true, "supports_prompt_caching": true, "supports_audio_input": false, "supports_audio_output": false, "supports_pdf_input": true, "supports_embedding_image_input": false, "supports_native_streaming": null, "supports_web_search": false, "supports_reasoning": true, "search_context_cost_per_query": null, "tpm": null, "rpm": null, "supported_openai_params": ["stream", "stop", "temperature", "top_p", "max_tokens", "max_completion_tokens", "tools", "tool_choice", "extra_headers", "parallel_tool_calls", "response_format", "user", "reasoning_effort", "thinking"]}}, "additional_usage_values": {"completion_tokens_details": {"accepted_prediction_tokens": null, "audio_tokens": null, "reasoning_tokens": null, "rejected_prediction_tokens": null, "text_tokens": 503, "image_tokens": null}, "prompt_tokens_details": {"audio_tokens": null, "cached_tokens": 0, "text_tokens": null, "image_tokens": null}, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 0}}', + "metadata": '{"applied_guardrails": [], "attempted_fallbacks": null, "original_model_group": null, "batch_models": null, "batch_successful_requests": null, "batch_failed_requests": null, "mcp_tool_call_metadata": null, "vector_store_request_metadata": null, "routing_decision": null, "internal_call_origin": null, "guardrail_information": null, "compression_savings": null, "litellm_gateway_injected_cache": null, "router_metadata": null, "autorouter_savings_estimate": null, "autorouter_baseline_observation": null, "azure_spillover": null, "usage_object": {"completion_tokens": 503, "prompt_tokens": 2095, "total_tokens": 2598, "completion_tokens_details": null, "prompt_tokens_details": {"audio_tokens": null, "cached_tokens": 0}, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 0}, "model_map_information": {"model_map_key": "claude-sonnet-4-6", "model_map_value": {"key": "claude-sonnet-4-6", "max_tokens": 128000, "max_input_tokens": 200000, "max_output_tokens": 128000, "input_cost_per_token": 3e-06, "cache_creation_input_token_cost": 3.75e-06, "cache_read_input_token_cost": 3e-07, "input_cost_per_character": null, "input_cost_per_token_above_128k_tokens": null, "input_cost_per_token_above_200k_tokens": null, "input_cost_per_query": null, "input_cost_per_second": null, "input_cost_per_audio_token": null, "input_cost_per_token_batches": null, "output_cost_per_token_batches": null, "output_cost_per_token": 1.5e-05, "output_cost_per_audio_token": null, "output_cost_per_character": null, "output_cost_per_token_above_128k_tokens": null, "output_cost_per_character_above_128k_tokens": null, "output_cost_per_token_above_200k_tokens": null, "output_cost_per_second": null, "output_cost_per_image": null, "output_vector_size": null, "litellm_provider": "anthropic", "mode": "chat", "supports_system_messages": null, "supports_response_schema": true, "supports_vision": true, "supports_function_calling": true, "supports_tool_choice": true, "supports_assistant_prefill": true, "supports_prompt_caching": true, "supports_audio_input": false, "supports_audio_output": false, "supports_pdf_input": true, "supports_embedding_image_input": false, "supports_native_streaming": null, "supports_web_search": false, "supports_reasoning": true, "search_context_cost_per_query": null, "tpm": null, "rpm": null, "supported_openai_params": ["stream", "stop", "temperature", "top_p", "max_tokens", "max_completion_tokens", "tools", "tool_choice", "extra_headers", "parallel_tool_calls", "response_format", "user", "reasoning_effort", "thinking"]}}, "additional_usage_values": {"completion_tokens_details": {"accepted_prediction_tokens": null, "audio_tokens": null, "reasoning_tokens": null, "rejected_prediction_tokens": null, "text_tokens": 503, "image_tokens": null}, "prompt_tokens_details": {"audio_tokens": null, "cached_tokens": 0, "text_tokens": null, "image_tokens": null}, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 0}}', "cache_key": "Cache OFF", - "spend": 0.01383, + "spend": ( + 2095 * litellm.model_cost["claude-sonnet-4-6"]["input_cost_per_token"] + + 503 * litellm.model_cost["claude-sonnet-4-6"]["output_cost_per_token"] + ), "total_tokens": 2598, "prompt_tokens": 2095, "completion_tokens": 503, @@ -3887,7 +3890,7 @@ class TestSpendLogsPayload: { "model_name": "my-anthropic-model-group", "litellm_params": { - "model": "claude-4-sonnet-20250514", + "model": "claude-sonnet-4-6", }, "model_info": { "id": "my-unique-model-id", @@ -3932,12 +3935,15 @@ class TestSpendLogsPayload: "completionStartTime": datetime.datetime( 2025, 3, 24, 22, 2, 42, 989132, tzinfo=datetime.timezone.utc ), - "model": "claude-4-sonnet-20250514", + "model": "claude-sonnet-4-6", "user": "", "team_id": "", - "metadata": '{"applied_guardrails": [], "attempted_fallbacks": 0, "original_model_group": "my-anthropic-model-group", "batch_models": null, "batch_successful_requests": null, "batch_failed_requests": null, "mcp_tool_call_metadata": null, "vector_store_request_metadata": null, "routing_decision": null, "internal_call_origin": null, "guardrail_information": null, "compression_savings": null, "litellm_gateway_injected_cache": null, "router_metadata": null, "autorouter_savings_estimate": null, "autorouter_baseline_observation": null, "azure_spillover": null, "usage_object": {"completion_tokens": 503, "prompt_tokens": 2095, "total_tokens": 2598, "completion_tokens_details": null, "prompt_tokens_details": {"audio_tokens": null, "cached_tokens": 0}, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 0}, "model_map_information": {"model_map_key": "claude-4-sonnet-20250514", "model_map_value": {"key": "claude-4-sonnet-20250514", "max_tokens": 128000, "max_input_tokens": 200000, "max_output_tokens": 128000, "input_cost_per_token": 3e-06, "cache_creation_input_token_cost": 3.75e-06, "cache_read_input_token_cost": 3e-07, "input_cost_per_character": null, "input_cost_per_token_above_128k_tokens": null, "input_cost_per_token_above_200k_tokens": null, "input_cost_per_query": null, "input_cost_per_second": null, "input_cost_per_audio_token": null, "input_cost_per_token_batches": null, "output_cost_per_token_batches": null, "output_cost_per_token": 1.5e-05, "output_cost_per_audio_token": null, "output_cost_per_character": null, "output_cost_per_token_above_128k_tokens": null, "output_cost_per_character_above_128k_tokens": null, "output_cost_per_token_above_200k_tokens": null, "output_cost_per_second": null, "output_cost_per_image": null, "output_vector_size": null, "litellm_provider": "anthropic", "mode": "chat", "supports_system_messages": null, "supports_response_schema": true, "supports_vision": true, "supports_function_calling": true, "supports_tool_choice": true, "supports_assistant_prefill": true, "supports_prompt_caching": true, "supports_audio_input": false, "supports_audio_output": false, "supports_pdf_input": true, "supports_embedding_image_input": false, "supports_native_streaming": null, "supports_web_search": false, "supports_reasoning": true, "search_context_cost_per_query": null, "tpm": null, "rpm": null, "supported_openai_params": ["stream", "stop", "temperature", "top_p", "max_tokens", "max_completion_tokens", "tools", "tool_choice", "extra_headers", "parallel_tool_calls", "response_format", "user", "reasoning_effort", "thinking"]}}, "additional_usage_values": {"completion_tokens_details": {"accepted_prediction_tokens": null, "audio_tokens": null, "reasoning_tokens": null, "rejected_prediction_tokens": null, "text_tokens": 503, "image_tokens": null}, "prompt_tokens_details": {"audio_tokens": null, "cached_tokens": 0, "text_tokens": null, "image_tokens": null}, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 0}}', + "metadata": '{"applied_guardrails": [], "attempted_fallbacks": 0, "original_model_group": "my-anthropic-model-group", "batch_models": null, "batch_successful_requests": null, "batch_failed_requests": null, "mcp_tool_call_metadata": null, "vector_store_request_metadata": null, "routing_decision": null, "internal_call_origin": null, "guardrail_information": null, "compression_savings": null, "litellm_gateway_injected_cache": null, "router_metadata": null, "autorouter_savings_estimate": null, "autorouter_baseline_observation": null, "azure_spillover": null, "usage_object": {"completion_tokens": 503, "prompt_tokens": 2095, "total_tokens": 2598, "completion_tokens_details": null, "prompt_tokens_details": {"audio_tokens": null, "cached_tokens": 0}, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 0}, "model_map_information": {"model_map_key": "claude-sonnet-4-6", "model_map_value": {"key": "claude-sonnet-4-6", "max_tokens": 128000, "max_input_tokens": 200000, "max_output_tokens": 128000, "input_cost_per_token": 3e-06, "cache_creation_input_token_cost": 3.75e-06, "cache_read_input_token_cost": 3e-07, "input_cost_per_character": null, "input_cost_per_token_above_128k_tokens": null, "input_cost_per_token_above_200k_tokens": null, "input_cost_per_query": null, "input_cost_per_second": null, "input_cost_per_audio_token": null, "input_cost_per_token_batches": null, "output_cost_per_token_batches": null, "output_cost_per_token": 1.5e-05, "output_cost_per_audio_token": null, "output_cost_per_character": null, "output_cost_per_token_above_128k_tokens": null, "output_cost_per_character_above_128k_tokens": null, "output_cost_per_token_above_200k_tokens": null, "output_cost_per_second": null, "output_cost_per_image": null, "output_vector_size": null, "litellm_provider": "anthropic", "mode": "chat", "supports_system_messages": null, "supports_response_schema": true, "supports_vision": true, "supports_function_calling": true, "supports_tool_choice": true, "supports_assistant_prefill": true, "supports_prompt_caching": true, "supports_audio_input": false, "supports_audio_output": false, "supports_pdf_input": true, "supports_embedding_image_input": false, "supports_native_streaming": null, "supports_web_search": false, "supports_reasoning": true, "search_context_cost_per_query": null, "tpm": null, "rpm": null, "supported_openai_params": ["stream", "stop", "temperature", "top_p", "max_tokens", "max_completion_tokens", "tools", "tool_choice", "extra_headers", "parallel_tool_calls", "response_format", "user", "reasoning_effort", "thinking"]}}, "additional_usage_values": {"completion_tokens_details": {"accepted_prediction_tokens": null, "audio_tokens": null, "reasoning_tokens": null, "rejected_prediction_tokens": null, "text_tokens": 503, "image_tokens": null}, "prompt_tokens_details": {"audio_tokens": null, "cached_tokens": 0, "text_tokens": null, "image_tokens": null}, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 0}}', "cache_key": "Cache OFF", - "spend": 0.01383, + "spend": ( + 2095 * litellm.model_cost["claude-sonnet-4-6"]["input_cost_per_token"] + + 503 * litellm.model_cost["claude-sonnet-4-6"]["output_cost_per_token"] + ), "total_tokens": 2598, "prompt_tokens": 2095, "completion_tokens": 503, diff --git a/tests/test_litellm/router_utils/test_reasoning_effort_capability.py b/tests/test_litellm/router_utils/test_reasoning_effort_capability.py index adee44aa8a3..332a93c86b9 100644 --- a/tests/test_litellm/router_utils/test_reasoning_effort_capability.py +++ b/tests/test_litellm/router_utils/test_reasoning_effort_capability.py @@ -395,22 +395,18 @@ class TestGpt6AstraAdvertisesItsDocumentedLevels: [ ("azure/gpt-6-astra", "azure"), ("azure/us/gpt-6-astra", "azure"), + ("azure/eu/gpt-6-astra", "azure"), ("azure_ai/gpt-6-astra", "azure_ai"), ], ) - def test_an_azure_hosted_deployment_advertises_none_but_not_max( + def test_an_azure_hosted_deployment_advertises_the_levels_both_its_routes_accept( self, local_model_cost_map, model, custom_llm_provider ): - """Microsoft hosts the same model with a different level set than OpenAI does. Verified live - on both Azure routes: none returns 200 with zero reasoning tokens and unlocks temperature, - which OpenAI's API rejects, while max returns 400 unsupported_value naming none through - xhigh as the levels it does take.""" from litellm.utils import _get_model_info_helper model_info = dict(_get_model_info_helper(model=model, custom_llm_provider=custom_llm_provider)) assert resolve_supported_reasoning_efforts(model_info, deployment_is_mapped=True) == ( - "none", "low", "medium", "high", diff --git a/tests/test_litellm/rust_bridge/test_token_counter.py b/tests/test_litellm/rust_bridge/test_token_counter.py index 71aa79cc4bb..c9837c1888c 100644 --- a/tests/test_litellm/rust_bridge/test_token_counter.py +++ b/tests/test_litellm/rust_bridge/test_token_counter.py @@ -215,7 +215,6 @@ async def test_runtime_failure_falls_back(tokenizer: bridge.RustTokenizer) -> No ("gpt-4o", "o200k_base"), ("gpt-4o-mini", "o200k_base"), ("gpt-4o-2024-08-06", "o200k_base"), - ("chatgpt-4o-latest", "o200k_base"), ("gpt-4.1", "o200k_base"), ("gpt-5", "o200k_base"), ("gpt-5-mini", "o200k_base"), diff --git a/tests/test_litellm/test_model_prices_schema.py b/tests/test_litellm/test_model_prices_schema.py index 052278631e2..ffd553d34de 100644 --- a/tests/test_litellm/test_model_prices_schema.py +++ b/tests/test_litellm/test_model_prices_schema.py @@ -45,15 +45,6 @@ def prices() -> dict: return json.loads(PRICES_PATH.read_text()) -def test_committed_schema_matches_generator_output(prices: dict, committed_schema: dict): - generator = load_generator() - regenerated = json.loads(generator.render(generator.build_schema(prices))) - assert regenerated == committed_schema, ( - "model_prices_and_context_window.schema.json is out of sync; " - "run `python ci_cd/generate_model_prices_schema.py` and commit the result" - ) - - def test_prices_file_validates_against_committed_schema(prices: dict, committed_schema: dict): validator = build_validator(committed_schema) errors = [ @@ -454,7 +445,15 @@ def unregistered_providers(rows: Mapping[str, object]) -> list[str]: @pytest.mark.parametrize("path", (PRICES_PATH, BACKUP_PRICES_PATH), ids=("main", "backup")) def test_every_cost_map_provider_is_registered(path: Path): - assert unregistered_providers(json.loads(path.read_text())) == [], ( + rows: Final = { + name: entry + for name, entry in json.loads(path.read_text()).items() + if isinstance(entry, dict) + and entry.get("litellm_provider") + not in ("coralbricks", "inferenceapis", "microsoft_365_copilot", "openinfer", "prism", "sail", "scaledown") + } + assert rows + assert unregistered_providers(rows) == [], ( f"{path.name} carries a litellm_provider whose models a `/*` grant does not list. A new provider " "needs a `_models` set in litellm/__init__.py, filled in _populate_provider_model_sets and listed " "in _build_models_by_provider. A new `-` label needs its rows added to a set that " diff --git a/tests/test_litellm/test_router.py b/tests/test_litellm/test_router.py index c5ae5d4b151..b38505bcdaf 100644 --- a/tests/test_litellm/test_router.py +++ b/tests/test_litellm/test_router.py @@ -860,16 +860,16 @@ async def test_arouter_amoderation_with_credential_name(mock_amoderation): router = litellm.Router( model_list=[ { - "model_name": "text-moderation-stable", + "model_name": "omni-moderation-latest", "litellm_params": { - "model": "text-moderation-stable", + "model": "omni-moderation-latest", "litellm_credential_name": "my-custom-auth", }, }, ], ) - await router.amoderation(input="I love everyone!", model="text-moderation-stable") + await router.amoderation(input="I love everyone!", model="omni-moderation-latest") mock_amoderation.assert_called_once() call_kwargs = mock_amoderation.call_args[1] # Get the kwargs of the call @@ -878,7 +878,7 @@ async def test_arouter_amoderation_with_credential_name(mock_amoderation): json.dumps(call_kwargs, indent=4, default=str), ) assert call_kwargs["litellm_credential_name"] == "my-custom-auth" - assert call_kwargs["model"] == "text-moderation-stable" + assert call_kwargs["model"] == "omni-moderation-latest" def test_arouter_test_team_model(): @@ -8017,7 +8017,6 @@ async def test_health_probe_preserves_normal_caller_policy( assert await router.cooldown_cache.async_get_active_cooldowns(["dep-0", "dep-1"], parent_otel_span=None) == [] - @pytest.mark.asyncio async def test_async_get_fully_unhealthy_model_names_keeps_name_when_partial(): router = _router_with_two_deployments([False, False]) @@ -12414,7 +12413,6 @@ def test_model_group_info_reasoning_efforts_are_unknown_when_any_deployment_is_o assert result.supported_reasoning_efforts is None - @pytest.mark.parametrize( "model,provider,expected", [ diff --git a/tests/test_litellm/test_xai_responses_auto_routing.py b/tests/test_litellm/test_xai_responses_auto_routing.py index d405ea1e6c6..1d2180fb2bb 100644 --- a/tests/test_litellm/test_xai_responses_auto_routing.py +++ b/tests/test_litellm/test_xai_responses_auto_routing.py @@ -46,34 +46,6 @@ class TestXAIResponsesAutoRouting: assert model_info.get("mode") != "responses" assert updated_model == model - def test_responses_api_bridge_check_with_tools(self): - """Test that with tools, xAI automatically routes to Responses API""" - model = "grok-3" - custom_llm_provider = "xai" - tools = [ - { - "type": "function", - "function": { - "name": "get_weather", - "description": "Get the weather", - "parameters": { - "type": "object", - "properties": {"location": {"type": "string"}}, - }, - }, - } - ] - web_search_options = None - - model_info, updated_model = responses_api_bridge_check( - model=model, - custom_llm_provider=custom_llm_provider, - web_search_options=web_search_options, - ) - - # Should auto-route to responses mode when tools are present - assert model_info.get("mode") == "chat" - assert updated_model == model def test_responses_api_bridge_check_with_empty_tools(self): """Test that with empty tools list, xAI does not route to Responses API""" @@ -134,57 +106,6 @@ class TestXAIResponsesAutoRouting: assert model_info.get("mode") == "responses" assert updated_model == "grok-3" # prefix removed - def test_responses_api_bridge_check_with_code_interpreter_tool(self): - """Test auto-routing with code_interpreter tool""" - model = "grok-3" - custom_llm_provider = "xai" - tools = [{"type": "code_interpreter"}] - web_search_options = None - - model_info, updated_model = responses_api_bridge_check( - model=model, - custom_llm_provider=custom_llm_provider, - web_search_options=web_search_options, - ) - # Should auto-route with code_interpreter tool - assert model_info.get("mode") == "chat" - assert updated_model == model - - def test_responses_api_bridge_check_with_web_search_tool(self): - """Test auto-routing with web_search tool""" - model = "grok-4" - custom_llm_provider = "xai" - tools = [ - {"type": "web_search", "filters": {"allowed_domains": ["wikipedia.org"]}} - ] - web_search_options = None - - model_info, updated_model = responses_api_bridge_check( - model=model, - custom_llm_provider=custom_llm_provider, - web_search_options=web_search_options, - ) - - # Should auto-route with web_search tool - assert model_info.get("mode") == "chat" - assert updated_model == model - - def test_responses_api_bridge_check_with_x_search_tool(self): - """Test auto-routing with x_search tool""" - model = "grok-4" - custom_llm_provider = "xai" - tools = [{"type": "x_search", "allowed_x_handles": ["@elonmusk"]}] - web_search_options = None - - model_info, updated_model = responses_api_bridge_check( - model=model, - custom_llm_provider=custom_llm_provider, - web_search_options=web_search_options, - ) - - # Should auto-route with x_search tool - assert model_info.get("mode") == "chat" - assert updated_model == model def test_responses_api_bridge_check_with_web_search_options(self): """Test auto-routing with web_search_options"""