From 134d1f3899ebb9f11415e1917c27d372f9e98960 Mon Sep 17 00:00:00 2001 From: mateo Date: Thu, 10 Sep 2026 13:17:14 +0000 Subject: [PATCH] fix(model_prices): registry audit 2026-09-10, absorb open pricing PRs Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com> --- ...odel_prices_and_context_window_backup.json | 60 +++++++++++++++--- model_prices_and_context_window.json | 60 +++++++++++++++--- .../llms/openai/test_gpt5_transformation.py | 63 +++++++++++-------- 3 files changed, 137 insertions(+), 46 deletions(-) diff --git a/litellm/model_prices_and_context_window_backup.json b/litellm/model_prices_and_context_window_backup.json index 0d2eda93323..d8160a3d842 100644 --- a/litellm/model_prices_and_context_window_backup.json +++ b/litellm/model_prices_and_context_window_backup.json @@ -30295,7 +30295,7 @@ "supports_none_reasoning_effort": true, "default_reasoning_effort": "none", "supports_xhigh_reasoning_effort": false, - "supports_minimal_reasoning_effort": true + "supports_minimal_reasoning_effort": false }, "gpt-5.1-2025-11-13": { "cache_read_input_token_cost": 1.25e-07, @@ -30340,7 +30340,7 @@ "supports_none_reasoning_effort": true, "default_reasoning_effort": "none", "supports_xhigh_reasoning_effort": false, - "supports_minimal_reasoning_effort": true + "supports_minimal_reasoning_effort": false }, "gpt-5.1-chat-latest": { "cache_read_input_token_cost": 1.25e-07, @@ -30432,7 +30432,7 @@ "supports_none_reasoning_effort": true, "default_reasoning_effort": "none", "supports_xhigh_reasoning_effort": true, - "supports_minimal_reasoning_effort": true + "supports_minimal_reasoning_effort": false }, "gpt-5.2-2025-12-11": { "cache_read_input_token_cost": 1.75e-07, @@ -30478,7 +30478,7 @@ "supports_none_reasoning_effort": true, "default_reasoning_effort": "none", "supports_xhigh_reasoning_effort": true, - "supports_minimal_reasoning_effort": true + "supports_minimal_reasoning_effort": false }, "gpt-5.2-chat-latest": { "cache_read_input_token_cost": 1.75e-07, @@ -31474,7 +31474,7 @@ "supports_none_reasoning_effort": true, "default_reasoning_effort": "none", "supports_xhigh_reasoning_effort": true, - "supports_minimal_reasoning_effort": true, + "supports_minimal_reasoning_effort": false, "input_cost_per_token_above_272k_tokens_flex": 2.5e-06, "output_cost_per_token_above_272k_tokens_flex": 1.125e-05, "cache_read_input_token_cost_above_272k_tokens_flex": 2.5e-07 @@ -31526,7 +31526,7 @@ "supports_none_reasoning_effort": true, "default_reasoning_effort": "none", "supports_xhigh_reasoning_effort": true, - "supports_minimal_reasoning_effort": true, + "supports_minimal_reasoning_effort": false, "input_cost_per_token_above_272k_tokens_flex": 2.5e-06, "output_cost_per_token_above_272k_tokens_flex": 1.125e-05, "cache_read_input_token_cost_above_272k_tokens_flex": 2.5e-07 @@ -31578,7 +31578,7 @@ "supports_web_search": true, "supports_none_reasoning_effort": false, "supports_xhigh_reasoning_effort": true, - "supports_minimal_reasoning_effort": true, + "supports_minimal_reasoning_effort": false, "input_cost_per_token_above_272k_tokens_flex": 3e-05, "output_cost_per_token_above_272k_tokens_flex": 0.000135 }, @@ -31629,7 +31629,7 @@ "supports_web_search": true, "supports_none_reasoning_effort": false, "supports_xhigh_reasoning_effort": true, - "supports_minimal_reasoning_effort": true, + "supports_minimal_reasoning_effort": false, "input_cost_per_token_above_272k_tokens_flex": 3e-05, "output_cost_per_token_above_272k_tokens_flex": 0.000135 }, @@ -33709,13 +33709,14 @@ "supports_tool_choice": true }, "jina-reranker-v2-base-multilingual": { - "input_cost_per_token": 1.8e-08, + "input_cost_per_token": 5e-08, "litellm_provider": "jina_ai", "max_input_tokens": 1024, "max_output_tokens": 1024, "max_tokens": 1024, "mode": "rerank", - "output_cost_per_token": 1.8e-08 + "output_cost_per_token": 0.0, + "source": "https://api.jina.ai/v1/models" }, "jp.anthropic.claude-sonnet-4-5-20250929-v1:0": { "cache_creation_input_token_cost": 4.125e-06, @@ -39969,6 +39970,45 @@ "supports_tool_choice": true, "supports_vision": true }, + "openrouter/openai/gpt-5.6-sol": { + "cache_creation_input_token_cost": 2.5e-06, + "cache_creation_input_token_cost_above_272k_tokens": 5e-06, + "cache_read_input_token_cost": 2e-07, + "cache_read_input_token_cost_above_272k_tokens": 4e-07, + "default_reasoning_effort": "medium", + "input_cost_per_token": 2e-06, + "input_cost_per_token_above_272k_tokens": 4e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_above_272k_tokens": 1.5e-05, + "reasoning_effort_levels": [ + "none", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "source": "https://openrouter.ai/openai/gpt-5.6-sol", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, "openrouter/openai/gpt-oss-120b": { "input_cost_per_token": 3.7e-08, "litellm_provider": "openrouter", diff --git a/model_prices_and_context_window.json b/model_prices_and_context_window.json index 0d2eda93323..d8160a3d842 100644 --- a/model_prices_and_context_window.json +++ b/model_prices_and_context_window.json @@ -30295,7 +30295,7 @@ "supports_none_reasoning_effort": true, "default_reasoning_effort": "none", "supports_xhigh_reasoning_effort": false, - "supports_minimal_reasoning_effort": true + "supports_minimal_reasoning_effort": false }, "gpt-5.1-2025-11-13": { "cache_read_input_token_cost": 1.25e-07, @@ -30340,7 +30340,7 @@ "supports_none_reasoning_effort": true, "default_reasoning_effort": "none", "supports_xhigh_reasoning_effort": false, - "supports_minimal_reasoning_effort": true + "supports_minimal_reasoning_effort": false }, "gpt-5.1-chat-latest": { "cache_read_input_token_cost": 1.25e-07, @@ -30432,7 +30432,7 @@ "supports_none_reasoning_effort": true, "default_reasoning_effort": "none", "supports_xhigh_reasoning_effort": true, - "supports_minimal_reasoning_effort": true + "supports_minimal_reasoning_effort": false }, "gpt-5.2-2025-12-11": { "cache_read_input_token_cost": 1.75e-07, @@ -30478,7 +30478,7 @@ "supports_none_reasoning_effort": true, "default_reasoning_effort": "none", "supports_xhigh_reasoning_effort": true, - "supports_minimal_reasoning_effort": true + "supports_minimal_reasoning_effort": false }, "gpt-5.2-chat-latest": { "cache_read_input_token_cost": 1.75e-07, @@ -31474,7 +31474,7 @@ "supports_none_reasoning_effort": true, "default_reasoning_effort": "none", "supports_xhigh_reasoning_effort": true, - "supports_minimal_reasoning_effort": true, + "supports_minimal_reasoning_effort": false, "input_cost_per_token_above_272k_tokens_flex": 2.5e-06, "output_cost_per_token_above_272k_tokens_flex": 1.125e-05, "cache_read_input_token_cost_above_272k_tokens_flex": 2.5e-07 @@ -31526,7 +31526,7 @@ "supports_none_reasoning_effort": true, "default_reasoning_effort": "none", "supports_xhigh_reasoning_effort": true, - "supports_minimal_reasoning_effort": true, + "supports_minimal_reasoning_effort": false, "input_cost_per_token_above_272k_tokens_flex": 2.5e-06, "output_cost_per_token_above_272k_tokens_flex": 1.125e-05, "cache_read_input_token_cost_above_272k_tokens_flex": 2.5e-07 @@ -31578,7 +31578,7 @@ "supports_web_search": true, "supports_none_reasoning_effort": false, "supports_xhigh_reasoning_effort": true, - "supports_minimal_reasoning_effort": true, + "supports_minimal_reasoning_effort": false, "input_cost_per_token_above_272k_tokens_flex": 3e-05, "output_cost_per_token_above_272k_tokens_flex": 0.000135 }, @@ -31629,7 +31629,7 @@ "supports_web_search": true, "supports_none_reasoning_effort": false, "supports_xhigh_reasoning_effort": true, - "supports_minimal_reasoning_effort": true, + "supports_minimal_reasoning_effort": false, "input_cost_per_token_above_272k_tokens_flex": 3e-05, "output_cost_per_token_above_272k_tokens_flex": 0.000135 }, @@ -33709,13 +33709,14 @@ "supports_tool_choice": true }, "jina-reranker-v2-base-multilingual": { - "input_cost_per_token": 1.8e-08, + "input_cost_per_token": 5e-08, "litellm_provider": "jina_ai", "max_input_tokens": 1024, "max_output_tokens": 1024, "max_tokens": 1024, "mode": "rerank", - "output_cost_per_token": 1.8e-08 + "output_cost_per_token": 0.0, + "source": "https://api.jina.ai/v1/models" }, "jp.anthropic.claude-sonnet-4-5-20250929-v1:0": { "cache_creation_input_token_cost": 4.125e-06, @@ -39969,6 +39970,45 @@ "supports_tool_choice": true, "supports_vision": true }, + "openrouter/openai/gpt-5.6-sol": { + "cache_creation_input_token_cost": 2.5e-06, + "cache_creation_input_token_cost_above_272k_tokens": 5e-06, + "cache_read_input_token_cost": 2e-07, + "cache_read_input_token_cost_above_272k_tokens": 4e-07, + "default_reasoning_effort": "medium", + "input_cost_per_token": 2e-06, + "input_cost_per_token_above_272k_tokens": 4e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_above_272k_tokens": 1.5e-05, + "reasoning_effort_levels": [ + "none", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "source": "https://openrouter.ai/openai/gpt-5.6-sol", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, "openrouter/openai/gpt-oss-120b": { "input_cost_per_token": 3.7e-08, "litellm_provider": "openrouter", diff --git a/tests/test_litellm/llms/openai/test_gpt5_transformation.py b/tests/test_litellm/llms/openai/test_gpt5_transformation.py index c86ce4df2ac..b538fad71a2 100644 --- a/tests/test_litellm/llms/openai/test_gpt5_transformation.py +++ b/tests/test_litellm/llms/openai/test_gpt5_transformation.py @@ -276,6 +276,17 @@ def test_gpt5_drops_reasoning_effort_xhigh_when_requested(config: OpenAIConfig): assert "reasoning_effort" not in params +def test_gpt5_1_gpt5_2_gpt5_4_drop_minimal_reasoning_effort(config: OpenAIConfig): + for model in ("gpt-5.1", "gpt-5.2", "gpt-5.4", "gpt-5.4-pro"): + params = config.map_openai_params( + non_default_params={"reasoning_effort": "minimal"}, + optional_params={}, + model=model, + drop_params=True, + ) + assert "reasoning_effort" not in params + + # GPT-5.1 temperature handling tests def test_gpt5_1_model_detection(gpt5_config: OpenAIGPT5Config): """Test that models supporting reasoning_effort='none' are correctly detected via model map.""" @@ -388,26 +399,26 @@ def test_gpt5_4_mini_allows_reasoning_effort_none(config: OpenAIConfig): assert params["reasoning_effort"] == "none" -def test_gpt5_4_allows_reasoning_effort_minimal(config: OpenAIConfig): - """gpt-5.4 supports reasoning_effort='minimal'.""" - params = config.map_openai_params( - non_default_params={"reasoning_effort": "minimal"}, - optional_params={}, - model="gpt-5.4", - drop_params=False, - ) - assert params["reasoning_effort"] == "minimal" +def test_gpt5_4_rejects_reasoning_effort_minimal(config: OpenAIConfig): + """gpt-5.4 rejects reasoning_effort='minimal'.""" + with pytest.raises(litellm.utils.UnsupportedParamsError): + config.map_openai_params( + non_default_params={"reasoning_effort": "minimal"}, + optional_params={}, + model="gpt-5.4", + drop_params=False, + ) -def test_gpt5_4_pro_allows_reasoning_effort_minimal(config: OpenAIConfig): - """gpt-5.4-pro supports reasoning_effort='minimal'.""" - params = config.map_openai_params( - non_default_params={"reasoning_effort": "minimal"}, - optional_params={}, - model="gpt-5.4-pro", - drop_params=False, - ) - assert params["reasoning_effort"] == "minimal" +def test_gpt5_4_pro_rejects_reasoning_effort_minimal(config: OpenAIConfig): + """gpt-5.4-pro rejects reasoning_effort='minimal'.""" + with pytest.raises(litellm.utils.UnsupportedParamsError): + config.map_openai_params( + non_default_params={"reasoning_effort": "minimal"}, + optional_params={}, + model="gpt-5.4-pro", + drop_params=False, + ) def test_gpt5_4_mini_rejects_reasoning_effort_minimal(config: OpenAIConfig): @@ -468,13 +479,13 @@ def test_gpt5_minimal_dict_triggers_validation(config: OpenAIConfig): def test_gpt5_minimal_dict_accepted_for_supported_model(config: OpenAIConfig): - """Dict with effort='minimal' passes through for gpt-5.4+.""" + """Dict with effort='minimal' passes through for gpt-5.""" params = config.map_openai_params( non_default_params={ "reasoning_effort": {"effort": "minimal", "summary": "detailed"} }, optional_params={}, - model="gpt-5.4", + model="gpt-5", drop_params=False, ) assert params["reasoning_effort"] == "minimal" @@ -482,8 +493,8 @@ def test_gpt5_minimal_dict_accepted_for_supported_model(config: OpenAIConfig): def test_gpt5_supports_reasoning_effort_level_minimal(gpt5_config: OpenAIGPT5Config): """Test that _supports_reasoning_effort_level correctly identifies minimal support.""" - assert gpt5_config._supports_reasoning_effort_level("gpt-5.4", "minimal") - assert gpt5_config._supports_reasoning_effort_level("gpt-5.4-pro", "minimal") + assert not gpt5_config._supports_reasoning_effort_level("gpt-5.4", "minimal") + assert not gpt5_config._supports_reasoning_effort_level("gpt-5.4-pro", "minimal") assert not gpt5_config._supports_reasoning_effort_level("gpt-5.4-mini", "minimal") assert not gpt5_config._supports_reasoning_effort_level("gpt-5.4-nano", "minimal") @@ -504,10 +515,10 @@ def test_gpt5_minimal_explicitly_disabled_check(gpt5_config: OpenAIGPT5Config): assert gpt5_config._is_reasoning_effort_level_explicitly_disabled( "openai/gpt-5.4-mini", "minimal" ) - assert not gpt5_config._is_reasoning_effort_level_explicitly_disabled( + assert gpt5_config._is_reasoning_effort_level_explicitly_disabled( "gpt-5.4", "minimal" ) - assert not gpt5_config._is_reasoning_effort_level_explicitly_disabled( + assert gpt5_config._is_reasoning_effort_level_explicitly_disabled( "gpt-5.4-pro", "minimal" ) @@ -523,8 +534,8 @@ def test_is_explicitly_disabled_factory_minimal(): assert _is_explicitly_disabled_factory("gpt-5.4-mini", None, key) assert _is_explicitly_disabled_factory("gpt-5.4-nano", None, key) assert _is_explicitly_disabled_factory("openai/gpt-5.4-mini", None, key) - assert not _is_explicitly_disabled_factory("gpt-5.4", None, key) - assert not _is_explicitly_disabled_factory("gpt-5.4-pro", None, key) + assert _is_explicitly_disabled_factory("gpt-5.4", None, key) + assert _is_explicitly_disabled_factory("gpt-5.4-pro", None, key) assert not _is_explicitly_disabled_factory("gpt-5.4-turbo-preview", None, key)