From b349b9bf5007612b0bc3000f72f5ac0f294339d0 Mon Sep 17 00:00:00 2001 From: mateo-berri <277851410+mateo-berri@users.noreply.github.com> Date: Wed, 26 Aug 2026 10:50:44 -0700 Subject: [PATCH 1/4] fix(pricing): add azure gpt-5.6 cache write rates and correct data zone priority --- ...odel_prices_and_context_window_backup.json | 88 ++++++++++++++----- model_prices_and_context_window.json | 88 ++++++++++++++----- tests/test_litellm/test_cost_calculator.py | 68 ++++++++++++++ 3 files changed, 196 insertions(+), 48 deletions(-) diff --git a/litellm/model_prices_and_context_window_backup.json b/litellm/model_prices_and_context_window_backup.json index 9ca8d9e1bac..3f1d84cae97 100644 --- a/litellm/model_prices_and_context_window_backup.json +++ b/litellm/model_prices_and_context_window_backup.json @@ -6592,6 +6592,10 @@ "supports_web_search": true }, "azure/gpt-5.6": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_272k_tokens": 1.25e-05, + "cache_creation_input_token_cost_priority": 1.25e-05, + "cache_creation_input_token_cost_above_272k_tokens_priority": 2.5e-05, "cache_read_input_token_cost": 5e-07, "cache_read_input_token_cost_above_272k_tokens": 1e-06, "cache_read_input_token_cost_priority": 1e-06, @@ -6642,6 +6646,10 @@ "supports_minimal_reasoning_effort": false }, "azure/gpt-5.6-sol": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_272k_tokens": 1.25e-05, + "cache_creation_input_token_cost_priority": 1.25e-05, + "cache_creation_input_token_cost_above_272k_tokens_priority": 2.5e-05, "cache_read_input_token_cost": 5e-07, "cache_read_input_token_cost_above_272k_tokens": 1e-06, "cache_read_input_token_cost_priority": 1e-06, @@ -6693,6 +6701,10 @@ "supports_minimal_reasoning_effort": false }, "azure/gpt-5.6-terra": { + "cache_creation_input_token_cost": 2.5e-06, + "cache_creation_input_token_cost_above_272k_tokens": 5e-06, + "cache_creation_input_token_cost_priority": 5e-06, + "cache_creation_input_token_cost_above_272k_tokens_priority": 1e-05, "cache_read_input_token_cost": 2e-07, "cache_read_input_token_cost_above_272k_tokens": 4e-07, "cache_read_input_token_cost_priority": 4e-07, @@ -6744,6 +6756,10 @@ "supports_minimal_reasoning_effort": false }, "azure/gpt-5.6-luna": { + "cache_creation_input_token_cost": 2.5e-07, + "cache_creation_input_token_cost_above_272k_tokens": 5e-07, + "cache_creation_input_token_cost_priority": 5e-07, + "cache_creation_input_token_cost_above_272k_tokens_priority": 1e-06, "cache_read_input_token_cost": 2e-08, "cache_read_input_token_cost_above_272k_tokens": 4e-08, "cache_read_input_token_cost_priority": 4e-08, @@ -6795,12 +6811,15 @@ "supports_minimal_reasoning_effort": false }, "azure/us/gpt-5.6": { + "cache_creation_input_token_cost": 6.875e-06, + "cache_creation_input_token_cost_above_272k_tokens": 1.375e-05, + "cache_creation_input_token_cost_priority": 1.375e-05, "cache_read_input_token_cost": 5.5e-07, "cache_read_input_token_cost_above_272k_tokens": 1.1e-06, - "cache_read_input_token_cost_priority": 1.375e-06, + "cache_read_input_token_cost_priority": 1.1e-06, "input_cost_per_token": 5.5e-06, "input_cost_per_token_above_272k_tokens": 1.1e-05, - "input_cost_per_token_priority": 1.375e-05, + "input_cost_per_token_priority": 1.1e-05, "litellm_provider": "azure", "max_input_tokens": 922000, "max_output_tokens": 128000, @@ -6808,7 +6827,7 @@ "mode": "chat", "output_cost_per_token": 3.3e-05, "output_cost_per_token_above_272k_tokens": 4.95e-05, - "output_cost_per_token_priority": 8.25e-05, + "output_cost_per_token_priority": 6.6e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, "search_context_size_low": 0.01, @@ -6842,13 +6861,16 @@ "supports_minimal_reasoning_effort": false }, "azure/us/gpt-5.6-sol": { + "cache_creation_input_token_cost": 6.875e-06, + "cache_creation_input_token_cost_above_272k_tokens": 1.375e-05, + "cache_creation_input_token_cost_priority": 1.375e-05, "cache_read_input_token_cost": 5.5e-07, "cache_read_input_token_cost_above_272k_tokens": 1.1e-06, - "cache_read_input_token_cost_priority": 1.375e-06, + "cache_read_input_token_cost_priority": 1.1e-06, "deprecation_date": "2028-01-11", "input_cost_per_token": 5.5e-06, "input_cost_per_token_above_272k_tokens": 1.1e-05, - "input_cost_per_token_priority": 1.375e-05, + "input_cost_per_token_priority": 1.1e-05, "litellm_provider": "azure", "max_input_tokens": 922000, "max_output_tokens": 128000, @@ -6856,7 +6878,7 @@ "mode": "chat", "output_cost_per_token": 3.3e-05, "output_cost_per_token_above_272k_tokens": 4.95e-05, - "output_cost_per_token_priority": 8.25e-05, + "output_cost_per_token_priority": 6.6e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, "search_context_size_low": 0.01, @@ -6890,13 +6912,16 @@ "supports_minimal_reasoning_effort": false }, "azure/us/gpt-5.6-terra": { + "cache_creation_input_token_cost": 2.75e-06, + "cache_creation_input_token_cost_above_272k_tokens": 5.5e-06, + "cache_creation_input_token_cost_priority": 5.5e-06, "cache_read_input_token_cost": 2.2e-07, "cache_read_input_token_cost_above_272k_tokens": 4.4e-07, - "cache_read_input_token_cost_priority": 5.5e-07, + "cache_read_input_token_cost_priority": 4.4e-07, "deprecation_date": "2028-01-11", "input_cost_per_token": 2.2e-06, "input_cost_per_token_above_272k_tokens": 4.4e-06, - "input_cost_per_token_priority": 5.5e-06, + "input_cost_per_token_priority": 4.4e-06, "litellm_provider": "azure", "max_input_tokens": 922000, "max_output_tokens": 128000, @@ -6904,7 +6929,7 @@ "mode": "chat", "output_cost_per_token": 1.32e-05, "output_cost_per_token_above_272k_tokens": 1.98e-05, - "output_cost_per_token_priority": 3.3e-05, + "output_cost_per_token_priority": 2.64e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, "search_context_size_low": 0.01, @@ -6938,13 +6963,16 @@ "supports_minimal_reasoning_effort": false }, "azure/us/gpt-5.6-luna": { + "cache_creation_input_token_cost": 2.75e-07, + "cache_creation_input_token_cost_above_272k_tokens": 5.5e-07, + "cache_creation_input_token_cost_priority": 5.5e-07, "cache_read_input_token_cost": 2.2e-08, "cache_read_input_token_cost_above_272k_tokens": 4.4e-08, - "cache_read_input_token_cost_priority": 5.5e-08, + "cache_read_input_token_cost_priority": 4.4e-08, "deprecation_date": "2028-01-11", "input_cost_per_token": 2.2e-07, "input_cost_per_token_above_272k_tokens": 4.4e-07, - "input_cost_per_token_priority": 5.5e-07, + "input_cost_per_token_priority": 4.4e-07, "litellm_provider": "azure", "max_input_tokens": 922000, "max_output_tokens": 128000, @@ -6952,7 +6980,7 @@ "mode": "chat", "output_cost_per_token": 1.32e-06, "output_cost_per_token_above_272k_tokens": 1.98e-06, - "output_cost_per_token_priority": 3.3e-06, + "output_cost_per_token_priority": 2.64e-06, "search_context_cost_per_query": { "search_context_size_high": 0.01, "search_context_size_low": 0.01, @@ -6986,12 +7014,15 @@ "supports_minimal_reasoning_effort": false }, "azure/eu/gpt-5.6": { + "cache_creation_input_token_cost": 6.875e-06, + "cache_creation_input_token_cost_above_272k_tokens": 1.375e-05, + "cache_creation_input_token_cost_priority": 1.375e-05, "cache_read_input_token_cost": 5.5e-07, "cache_read_input_token_cost_above_272k_tokens": 1.1e-06, - "cache_read_input_token_cost_priority": 1.375e-06, + "cache_read_input_token_cost_priority": 1.1e-06, "input_cost_per_token": 5.5e-06, "input_cost_per_token_above_272k_tokens": 1.1e-05, - "input_cost_per_token_priority": 1.375e-05, + "input_cost_per_token_priority": 1.1e-05, "litellm_provider": "azure", "max_input_tokens": 922000, "max_output_tokens": 128000, @@ -6999,7 +7030,7 @@ "mode": "chat", "output_cost_per_token": 3.3e-05, "output_cost_per_token_above_272k_tokens": 4.95e-05, - "output_cost_per_token_priority": 8.25e-05, + "output_cost_per_token_priority": 6.6e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, "search_context_size_low": 0.01, @@ -7033,13 +7064,16 @@ "supports_minimal_reasoning_effort": false }, "azure/eu/gpt-5.6-sol": { + "cache_creation_input_token_cost": 6.875e-06, + "cache_creation_input_token_cost_above_272k_tokens": 1.375e-05, + "cache_creation_input_token_cost_priority": 1.375e-05, "cache_read_input_token_cost": 5.5e-07, "cache_read_input_token_cost_above_272k_tokens": 1.1e-06, - "cache_read_input_token_cost_priority": 1.375e-06, + "cache_read_input_token_cost_priority": 1.1e-06, "deprecation_date": "2028-01-11", "input_cost_per_token": 5.5e-06, "input_cost_per_token_above_272k_tokens": 1.1e-05, - "input_cost_per_token_priority": 1.375e-05, + "input_cost_per_token_priority": 1.1e-05, "litellm_provider": "azure", "max_input_tokens": 922000, "max_output_tokens": 128000, @@ -7047,7 +7081,7 @@ "mode": "chat", "output_cost_per_token": 3.3e-05, "output_cost_per_token_above_272k_tokens": 4.95e-05, - "output_cost_per_token_priority": 8.25e-05, + "output_cost_per_token_priority": 6.6e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, "search_context_size_low": 0.01, @@ -7081,13 +7115,16 @@ "supports_minimal_reasoning_effort": false }, "azure/eu/gpt-5.6-terra": { + "cache_creation_input_token_cost": 2.75e-06, + "cache_creation_input_token_cost_above_272k_tokens": 5.5e-06, + "cache_creation_input_token_cost_priority": 5.5e-06, "cache_read_input_token_cost": 2.2e-07, "cache_read_input_token_cost_above_272k_tokens": 4.4e-07, - "cache_read_input_token_cost_priority": 5.5e-07, + "cache_read_input_token_cost_priority": 4.4e-07, "deprecation_date": "2028-01-11", "input_cost_per_token": 2.2e-06, "input_cost_per_token_above_272k_tokens": 4.4e-06, - "input_cost_per_token_priority": 5.5e-06, + "input_cost_per_token_priority": 4.4e-06, "litellm_provider": "azure", "max_input_tokens": 922000, "max_output_tokens": 128000, @@ -7095,7 +7132,7 @@ "mode": "chat", "output_cost_per_token": 1.32e-05, "output_cost_per_token_above_272k_tokens": 1.98e-05, - "output_cost_per_token_priority": 3.3e-05, + "output_cost_per_token_priority": 2.64e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, "search_context_size_low": 0.01, @@ -7129,13 +7166,16 @@ "supports_minimal_reasoning_effort": false }, "azure/eu/gpt-5.6-luna": { + "cache_creation_input_token_cost": 2.75e-07, + "cache_creation_input_token_cost_above_272k_tokens": 5.5e-07, + "cache_creation_input_token_cost_priority": 5.5e-07, "cache_read_input_token_cost": 2.2e-08, "cache_read_input_token_cost_above_272k_tokens": 4.4e-08, - "cache_read_input_token_cost_priority": 5.5e-08, + "cache_read_input_token_cost_priority": 4.4e-08, "deprecation_date": "2028-01-11", "input_cost_per_token": 2.2e-07, "input_cost_per_token_above_272k_tokens": 4.4e-07, - "input_cost_per_token_priority": 5.5e-07, + "input_cost_per_token_priority": 4.4e-07, "litellm_provider": "azure", "max_input_tokens": 922000, "max_output_tokens": 128000, @@ -7143,7 +7183,7 @@ "mode": "chat", "output_cost_per_token": 1.32e-06, "output_cost_per_token_above_272k_tokens": 1.98e-06, - "output_cost_per_token_priority": 3.3e-06, + "output_cost_per_token_priority": 2.64e-06, "search_context_cost_per_query": { "search_context_size_high": 0.01, "search_context_size_low": 0.01, diff --git a/model_prices_and_context_window.json b/model_prices_and_context_window.json index 9ca8d9e1bac..3f1d84cae97 100644 --- a/model_prices_and_context_window.json +++ b/model_prices_and_context_window.json @@ -6592,6 +6592,10 @@ "supports_web_search": true }, "azure/gpt-5.6": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_272k_tokens": 1.25e-05, + "cache_creation_input_token_cost_priority": 1.25e-05, + "cache_creation_input_token_cost_above_272k_tokens_priority": 2.5e-05, "cache_read_input_token_cost": 5e-07, "cache_read_input_token_cost_above_272k_tokens": 1e-06, "cache_read_input_token_cost_priority": 1e-06, @@ -6642,6 +6646,10 @@ "supports_minimal_reasoning_effort": false }, "azure/gpt-5.6-sol": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_272k_tokens": 1.25e-05, + "cache_creation_input_token_cost_priority": 1.25e-05, + "cache_creation_input_token_cost_above_272k_tokens_priority": 2.5e-05, "cache_read_input_token_cost": 5e-07, "cache_read_input_token_cost_above_272k_tokens": 1e-06, "cache_read_input_token_cost_priority": 1e-06, @@ -6693,6 +6701,10 @@ "supports_minimal_reasoning_effort": false }, "azure/gpt-5.6-terra": { + "cache_creation_input_token_cost": 2.5e-06, + "cache_creation_input_token_cost_above_272k_tokens": 5e-06, + "cache_creation_input_token_cost_priority": 5e-06, + "cache_creation_input_token_cost_above_272k_tokens_priority": 1e-05, "cache_read_input_token_cost": 2e-07, "cache_read_input_token_cost_above_272k_tokens": 4e-07, "cache_read_input_token_cost_priority": 4e-07, @@ -6744,6 +6756,10 @@ "supports_minimal_reasoning_effort": false }, "azure/gpt-5.6-luna": { + "cache_creation_input_token_cost": 2.5e-07, + "cache_creation_input_token_cost_above_272k_tokens": 5e-07, + "cache_creation_input_token_cost_priority": 5e-07, + "cache_creation_input_token_cost_above_272k_tokens_priority": 1e-06, "cache_read_input_token_cost": 2e-08, "cache_read_input_token_cost_above_272k_tokens": 4e-08, "cache_read_input_token_cost_priority": 4e-08, @@ -6795,12 +6811,15 @@ "supports_minimal_reasoning_effort": false }, "azure/us/gpt-5.6": { + "cache_creation_input_token_cost": 6.875e-06, + "cache_creation_input_token_cost_above_272k_tokens": 1.375e-05, + "cache_creation_input_token_cost_priority": 1.375e-05, "cache_read_input_token_cost": 5.5e-07, "cache_read_input_token_cost_above_272k_tokens": 1.1e-06, - "cache_read_input_token_cost_priority": 1.375e-06, + "cache_read_input_token_cost_priority": 1.1e-06, "input_cost_per_token": 5.5e-06, "input_cost_per_token_above_272k_tokens": 1.1e-05, - "input_cost_per_token_priority": 1.375e-05, + "input_cost_per_token_priority": 1.1e-05, "litellm_provider": "azure", "max_input_tokens": 922000, "max_output_tokens": 128000, @@ -6808,7 +6827,7 @@ "mode": "chat", "output_cost_per_token": 3.3e-05, "output_cost_per_token_above_272k_tokens": 4.95e-05, - "output_cost_per_token_priority": 8.25e-05, + "output_cost_per_token_priority": 6.6e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, "search_context_size_low": 0.01, @@ -6842,13 +6861,16 @@ "supports_minimal_reasoning_effort": false }, "azure/us/gpt-5.6-sol": { + "cache_creation_input_token_cost": 6.875e-06, + "cache_creation_input_token_cost_above_272k_tokens": 1.375e-05, + "cache_creation_input_token_cost_priority": 1.375e-05, "cache_read_input_token_cost": 5.5e-07, "cache_read_input_token_cost_above_272k_tokens": 1.1e-06, - "cache_read_input_token_cost_priority": 1.375e-06, + "cache_read_input_token_cost_priority": 1.1e-06, "deprecation_date": "2028-01-11", "input_cost_per_token": 5.5e-06, "input_cost_per_token_above_272k_tokens": 1.1e-05, - "input_cost_per_token_priority": 1.375e-05, + "input_cost_per_token_priority": 1.1e-05, "litellm_provider": "azure", "max_input_tokens": 922000, "max_output_tokens": 128000, @@ -6856,7 +6878,7 @@ "mode": "chat", "output_cost_per_token": 3.3e-05, "output_cost_per_token_above_272k_tokens": 4.95e-05, - "output_cost_per_token_priority": 8.25e-05, + "output_cost_per_token_priority": 6.6e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, "search_context_size_low": 0.01, @@ -6890,13 +6912,16 @@ "supports_minimal_reasoning_effort": false }, "azure/us/gpt-5.6-terra": { + "cache_creation_input_token_cost": 2.75e-06, + "cache_creation_input_token_cost_above_272k_tokens": 5.5e-06, + "cache_creation_input_token_cost_priority": 5.5e-06, "cache_read_input_token_cost": 2.2e-07, "cache_read_input_token_cost_above_272k_tokens": 4.4e-07, - "cache_read_input_token_cost_priority": 5.5e-07, + "cache_read_input_token_cost_priority": 4.4e-07, "deprecation_date": "2028-01-11", "input_cost_per_token": 2.2e-06, "input_cost_per_token_above_272k_tokens": 4.4e-06, - "input_cost_per_token_priority": 5.5e-06, + "input_cost_per_token_priority": 4.4e-06, "litellm_provider": "azure", "max_input_tokens": 922000, "max_output_tokens": 128000, @@ -6904,7 +6929,7 @@ "mode": "chat", "output_cost_per_token": 1.32e-05, "output_cost_per_token_above_272k_tokens": 1.98e-05, - "output_cost_per_token_priority": 3.3e-05, + "output_cost_per_token_priority": 2.64e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, "search_context_size_low": 0.01, @@ -6938,13 +6963,16 @@ "supports_minimal_reasoning_effort": false }, "azure/us/gpt-5.6-luna": { + "cache_creation_input_token_cost": 2.75e-07, + "cache_creation_input_token_cost_above_272k_tokens": 5.5e-07, + "cache_creation_input_token_cost_priority": 5.5e-07, "cache_read_input_token_cost": 2.2e-08, "cache_read_input_token_cost_above_272k_tokens": 4.4e-08, - "cache_read_input_token_cost_priority": 5.5e-08, + "cache_read_input_token_cost_priority": 4.4e-08, "deprecation_date": "2028-01-11", "input_cost_per_token": 2.2e-07, "input_cost_per_token_above_272k_tokens": 4.4e-07, - "input_cost_per_token_priority": 5.5e-07, + "input_cost_per_token_priority": 4.4e-07, "litellm_provider": "azure", "max_input_tokens": 922000, "max_output_tokens": 128000, @@ -6952,7 +6980,7 @@ "mode": "chat", "output_cost_per_token": 1.32e-06, "output_cost_per_token_above_272k_tokens": 1.98e-06, - "output_cost_per_token_priority": 3.3e-06, + "output_cost_per_token_priority": 2.64e-06, "search_context_cost_per_query": { "search_context_size_high": 0.01, "search_context_size_low": 0.01, @@ -6986,12 +7014,15 @@ "supports_minimal_reasoning_effort": false }, "azure/eu/gpt-5.6": { + "cache_creation_input_token_cost": 6.875e-06, + "cache_creation_input_token_cost_above_272k_tokens": 1.375e-05, + "cache_creation_input_token_cost_priority": 1.375e-05, "cache_read_input_token_cost": 5.5e-07, "cache_read_input_token_cost_above_272k_tokens": 1.1e-06, - "cache_read_input_token_cost_priority": 1.375e-06, + "cache_read_input_token_cost_priority": 1.1e-06, "input_cost_per_token": 5.5e-06, "input_cost_per_token_above_272k_tokens": 1.1e-05, - "input_cost_per_token_priority": 1.375e-05, + "input_cost_per_token_priority": 1.1e-05, "litellm_provider": "azure", "max_input_tokens": 922000, "max_output_tokens": 128000, @@ -6999,7 +7030,7 @@ "mode": "chat", "output_cost_per_token": 3.3e-05, "output_cost_per_token_above_272k_tokens": 4.95e-05, - "output_cost_per_token_priority": 8.25e-05, + "output_cost_per_token_priority": 6.6e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, "search_context_size_low": 0.01, @@ -7033,13 +7064,16 @@ "supports_minimal_reasoning_effort": false }, "azure/eu/gpt-5.6-sol": { + "cache_creation_input_token_cost": 6.875e-06, + "cache_creation_input_token_cost_above_272k_tokens": 1.375e-05, + "cache_creation_input_token_cost_priority": 1.375e-05, "cache_read_input_token_cost": 5.5e-07, "cache_read_input_token_cost_above_272k_tokens": 1.1e-06, - "cache_read_input_token_cost_priority": 1.375e-06, + "cache_read_input_token_cost_priority": 1.1e-06, "deprecation_date": "2028-01-11", "input_cost_per_token": 5.5e-06, "input_cost_per_token_above_272k_tokens": 1.1e-05, - "input_cost_per_token_priority": 1.375e-05, + "input_cost_per_token_priority": 1.1e-05, "litellm_provider": "azure", "max_input_tokens": 922000, "max_output_tokens": 128000, @@ -7047,7 +7081,7 @@ "mode": "chat", "output_cost_per_token": 3.3e-05, "output_cost_per_token_above_272k_tokens": 4.95e-05, - "output_cost_per_token_priority": 8.25e-05, + "output_cost_per_token_priority": 6.6e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, "search_context_size_low": 0.01, @@ -7081,13 +7115,16 @@ "supports_minimal_reasoning_effort": false }, "azure/eu/gpt-5.6-terra": { + "cache_creation_input_token_cost": 2.75e-06, + "cache_creation_input_token_cost_above_272k_tokens": 5.5e-06, + "cache_creation_input_token_cost_priority": 5.5e-06, "cache_read_input_token_cost": 2.2e-07, "cache_read_input_token_cost_above_272k_tokens": 4.4e-07, - "cache_read_input_token_cost_priority": 5.5e-07, + "cache_read_input_token_cost_priority": 4.4e-07, "deprecation_date": "2028-01-11", "input_cost_per_token": 2.2e-06, "input_cost_per_token_above_272k_tokens": 4.4e-06, - "input_cost_per_token_priority": 5.5e-06, + "input_cost_per_token_priority": 4.4e-06, "litellm_provider": "azure", "max_input_tokens": 922000, "max_output_tokens": 128000, @@ -7095,7 +7132,7 @@ "mode": "chat", "output_cost_per_token": 1.32e-05, "output_cost_per_token_above_272k_tokens": 1.98e-05, - "output_cost_per_token_priority": 3.3e-05, + "output_cost_per_token_priority": 2.64e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, "search_context_size_low": 0.01, @@ -7129,13 +7166,16 @@ "supports_minimal_reasoning_effort": false }, "azure/eu/gpt-5.6-luna": { + "cache_creation_input_token_cost": 2.75e-07, + "cache_creation_input_token_cost_above_272k_tokens": 5.5e-07, + "cache_creation_input_token_cost_priority": 5.5e-07, "cache_read_input_token_cost": 2.2e-08, "cache_read_input_token_cost_above_272k_tokens": 4.4e-08, - "cache_read_input_token_cost_priority": 5.5e-08, + "cache_read_input_token_cost_priority": 4.4e-08, "deprecation_date": "2028-01-11", "input_cost_per_token": 2.2e-07, "input_cost_per_token_above_272k_tokens": 4.4e-07, - "input_cost_per_token_priority": 5.5e-07, + "input_cost_per_token_priority": 4.4e-07, "litellm_provider": "azure", "max_input_tokens": 922000, "max_output_tokens": 128000, @@ -7143,7 +7183,7 @@ "mode": "chat", "output_cost_per_token": 1.32e-06, "output_cost_per_token_above_272k_tokens": 1.98e-06, - "output_cost_per_token_priority": 3.3e-06, + "output_cost_per_token_priority": 2.64e-06, "search_context_cost_per_query": { "search_context_size_high": 0.01, "search_context_size_low": 0.01, diff --git a/tests/test_litellm/test_cost_calculator.py b/tests/test_litellm/test_cost_calculator.py index 1f1c9be973f..cd58c780710 100644 --- a/tests/test_litellm/test_cost_calculator.py +++ b/tests/test_litellm/test_cost_calculator.py @@ -1718,6 +1718,74 @@ def test_azure_ai_cache_cost_calculation(_local_model_cost_map): ), f"Output cost mismatch: got {output_cost}, expected {expected_output_cost}" + +AZURE_GPT_5_6_MAP_KEYS = ( + "azure/gpt-5.6", + "azure/gpt-5.6-sol", + "azure/gpt-5.6-terra", + "azure/gpt-5.6-luna", + "azure/us/gpt-5.6", + "azure/us/gpt-5.6-sol", + "azure/us/gpt-5.6-terra", + "azure/us/gpt-5.6-luna", + "azure/eu/gpt-5.6", + "azure/eu/gpt-5.6-sol", + "azure/eu/gpt-5.6-terra", + "azure/eu/gpt-5.6-luna", +) + + +def test_azure_gpt_5_6_cache_write_tokens_are_billed(_local_model_cost_map): + """ + Azure bills gpt-5.6 prompt cache writes at 1.25x the input rate on every + tier, but the azure entries carried no ``cache_creation_input_token_cost``, + so cache-write tokens were billed at the plain input rate instead. + """ + from litellm.litellm_core_utils.llm_cost_calc.utils import generic_cost_per_token + from litellm.types.utils import PromptTokensDetailsWrapper, Usage + + usage = Usage( + completion_tokens=100, + prompt_tokens=2000, + total_tokens=2100, + prompt_tokens_details=PromptTokensDetailsWrapper(cached_tokens=0, text_tokens=687), + cache_creation_input_tokens=1313, + ) + + input_cost, output_cost = generic_cost_per_token( + model="azure/gpt-5.6-luna", usage=usage, custom_llm_provider="azure" + ) + + assert input_cost == pytest.approx(687 * 2e-07 + 1313 * 2.5e-07) + assert output_cost == pytest.approx(100 * 1.2e-06) + + +@pytest.mark.parametrize("model", AZURE_GPT_5_6_MAP_KEYS) +def test_azure_gpt_5_6_rates_match_azure_price_page(_local_model_cost_map, model): + """ + Per the Azure OpenAI price page (rendered 2026-08-26): cache writes cost + 1.25x input on every gpt-5.6 tier, and Data Zone costs 1.1x Global for + standard and priority alike (us/eu priority rates previously sat at 1.25x). + """ + entry = litellm.model_cost[model] + input_keys = [key for key in entry if key.startswith("input_cost_per_token")] + assert input_keys + for key in input_keys: + suffix = key[len("input_cost_per_token") :] + assert entry["cache_creation_input_token_cost" + suffix] == pytest.approx(entry[key] * 1.25) + + zone = model.split("/")[1] + if zone in ("us", "eu"): + global_entry = litellm.model_cost["azure/" + model.split("/", 2)[2]] + token_cost_keys = [ + key + for key in entry + if key.startswith(("input_cost_per_token", "output_cost_per_token", "cache_read", "cache_creation")) + ] + assert len(token_cost_keys) >= 9 + for key in token_cost_keys: + assert entry[key] == pytest.approx(global_entry[key] * 1.1), key + def test_vertex_regional_deployment_costs_uplift_over_global(monkeypatch): """ Regression for https://github.com/BerriAI/litellm/issues/34393: two Vertex From 8d750a24684966aca08a2c6d4ae37c6cc4ccecea Mon Sep 17 00:00:00 2001 From: mateo-berri <277851410+mateo-berri@users.noreply.github.com> Date: Wed, 26 Aug 2026 11:31:53 -0700 Subject: [PATCH 2/4] chore(model_prices): regenerate schema for priority cache write tier keys --- model_prices_and_context_window.schema.json | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/model_prices_and_context_window.schema.json b/model_prices_and_context_window.schema.json index f7f60c7666d..4eee9d52bc7 100644 --- a/model_prices_and_context_window.schema.json +++ b/model_prices_and_context_window.schema.json @@ -104,6 +104,11 @@ "minimum": 0, "description": "Flex service-tier rate for the same-named base field." }, + "cache_creation_input_token_cost_above_272k_tokens_priority": { + "type": "number", + "minimum": 0, + "description": "Priority service-tier rate for the same-named base field." + }, "cache_creation_input_token_cost_flex": { "type": "number", "minimum": 0, From e97a84afcf9bce53f40a73430a08b94cf4fb62f9 Mon Sep 17 00:00:00 2001 From: mateo-berri <277851410+mateo-berri@users.noreply.github.com> Date: Wed, 26 Aug 2026 11:40:19 -0700 Subject: [PATCH 3/4] fix(model_prices): add above-272k priority rates to azure us/eu gpt-5.6 entries --- ...odel_prices_and_context_window_backup.json | 32 +++++++++++++++++++ model_prices_and_context_window.json | 32 +++++++++++++++++++ tests/test_litellm/test_cost_calculator.py | 9 +++--- 3 files changed, 68 insertions(+), 5 deletions(-) diff --git a/litellm/model_prices_and_context_window_backup.json b/litellm/model_prices_and_context_window_backup.json index 3f1d84cae97..57facc96319 100644 --- a/litellm/model_prices_and_context_window_backup.json +++ b/litellm/model_prices_and_context_window_backup.json @@ -6813,12 +6813,15 @@ "azure/us/gpt-5.6": { "cache_creation_input_token_cost": 6.875e-06, "cache_creation_input_token_cost_above_272k_tokens": 1.375e-05, + "cache_creation_input_token_cost_above_272k_tokens_priority": 2.75e-05, "cache_creation_input_token_cost_priority": 1.375e-05, "cache_read_input_token_cost": 5.5e-07, "cache_read_input_token_cost_above_272k_tokens": 1.1e-06, + "cache_read_input_token_cost_above_272k_tokens_priority": 2.2e-06, "cache_read_input_token_cost_priority": 1.1e-06, "input_cost_per_token": 5.5e-06, "input_cost_per_token_above_272k_tokens": 1.1e-05, + "input_cost_per_token_above_272k_tokens_priority": 2.2e-05, "input_cost_per_token_priority": 1.1e-05, "litellm_provider": "azure", "max_input_tokens": 922000, @@ -6827,6 +6830,7 @@ "mode": "chat", "output_cost_per_token": 3.3e-05, "output_cost_per_token_above_272k_tokens": 4.95e-05, + "output_cost_per_token_above_272k_tokens_priority": 9.9e-05, "output_cost_per_token_priority": 6.6e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, @@ -6863,13 +6867,16 @@ "azure/us/gpt-5.6-sol": { "cache_creation_input_token_cost": 6.875e-06, "cache_creation_input_token_cost_above_272k_tokens": 1.375e-05, + "cache_creation_input_token_cost_above_272k_tokens_priority": 2.75e-05, "cache_creation_input_token_cost_priority": 1.375e-05, "cache_read_input_token_cost": 5.5e-07, "cache_read_input_token_cost_above_272k_tokens": 1.1e-06, + "cache_read_input_token_cost_above_272k_tokens_priority": 2.2e-06, "cache_read_input_token_cost_priority": 1.1e-06, "deprecation_date": "2028-01-11", "input_cost_per_token": 5.5e-06, "input_cost_per_token_above_272k_tokens": 1.1e-05, + "input_cost_per_token_above_272k_tokens_priority": 2.2e-05, "input_cost_per_token_priority": 1.1e-05, "litellm_provider": "azure", "max_input_tokens": 922000, @@ -6878,6 +6885,7 @@ "mode": "chat", "output_cost_per_token": 3.3e-05, "output_cost_per_token_above_272k_tokens": 4.95e-05, + "output_cost_per_token_above_272k_tokens_priority": 9.9e-05, "output_cost_per_token_priority": 6.6e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, @@ -6914,13 +6922,16 @@ "azure/us/gpt-5.6-terra": { "cache_creation_input_token_cost": 2.75e-06, "cache_creation_input_token_cost_above_272k_tokens": 5.5e-06, + "cache_creation_input_token_cost_above_272k_tokens_priority": 1.1e-05, "cache_creation_input_token_cost_priority": 5.5e-06, "cache_read_input_token_cost": 2.2e-07, "cache_read_input_token_cost_above_272k_tokens": 4.4e-07, + "cache_read_input_token_cost_above_272k_tokens_priority": 8.8e-07, "cache_read_input_token_cost_priority": 4.4e-07, "deprecation_date": "2028-01-11", "input_cost_per_token": 2.2e-06, "input_cost_per_token_above_272k_tokens": 4.4e-06, + "input_cost_per_token_above_272k_tokens_priority": 8.8e-06, "input_cost_per_token_priority": 4.4e-06, "litellm_provider": "azure", "max_input_tokens": 922000, @@ -6929,6 +6940,7 @@ "mode": "chat", "output_cost_per_token": 1.32e-05, "output_cost_per_token_above_272k_tokens": 1.98e-05, + "output_cost_per_token_above_272k_tokens_priority": 3.96e-05, "output_cost_per_token_priority": 2.64e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, @@ -6965,13 +6977,16 @@ "azure/us/gpt-5.6-luna": { "cache_creation_input_token_cost": 2.75e-07, "cache_creation_input_token_cost_above_272k_tokens": 5.5e-07, + "cache_creation_input_token_cost_above_272k_tokens_priority": 1.1e-06, "cache_creation_input_token_cost_priority": 5.5e-07, "cache_read_input_token_cost": 2.2e-08, "cache_read_input_token_cost_above_272k_tokens": 4.4e-08, + "cache_read_input_token_cost_above_272k_tokens_priority": 8.8e-08, "cache_read_input_token_cost_priority": 4.4e-08, "deprecation_date": "2028-01-11", "input_cost_per_token": 2.2e-07, "input_cost_per_token_above_272k_tokens": 4.4e-07, + "input_cost_per_token_above_272k_tokens_priority": 8.8e-07, "input_cost_per_token_priority": 4.4e-07, "litellm_provider": "azure", "max_input_tokens": 922000, @@ -6980,6 +6995,7 @@ "mode": "chat", "output_cost_per_token": 1.32e-06, "output_cost_per_token_above_272k_tokens": 1.98e-06, + "output_cost_per_token_above_272k_tokens_priority": 3.96e-06, "output_cost_per_token_priority": 2.64e-06, "search_context_cost_per_query": { "search_context_size_high": 0.01, @@ -7016,12 +7032,15 @@ "azure/eu/gpt-5.6": { "cache_creation_input_token_cost": 6.875e-06, "cache_creation_input_token_cost_above_272k_tokens": 1.375e-05, + "cache_creation_input_token_cost_above_272k_tokens_priority": 2.75e-05, "cache_creation_input_token_cost_priority": 1.375e-05, "cache_read_input_token_cost": 5.5e-07, "cache_read_input_token_cost_above_272k_tokens": 1.1e-06, + "cache_read_input_token_cost_above_272k_tokens_priority": 2.2e-06, "cache_read_input_token_cost_priority": 1.1e-06, "input_cost_per_token": 5.5e-06, "input_cost_per_token_above_272k_tokens": 1.1e-05, + "input_cost_per_token_above_272k_tokens_priority": 2.2e-05, "input_cost_per_token_priority": 1.1e-05, "litellm_provider": "azure", "max_input_tokens": 922000, @@ -7030,6 +7049,7 @@ "mode": "chat", "output_cost_per_token": 3.3e-05, "output_cost_per_token_above_272k_tokens": 4.95e-05, + "output_cost_per_token_above_272k_tokens_priority": 9.9e-05, "output_cost_per_token_priority": 6.6e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, @@ -7066,13 +7086,16 @@ "azure/eu/gpt-5.6-sol": { "cache_creation_input_token_cost": 6.875e-06, "cache_creation_input_token_cost_above_272k_tokens": 1.375e-05, + "cache_creation_input_token_cost_above_272k_tokens_priority": 2.75e-05, "cache_creation_input_token_cost_priority": 1.375e-05, "cache_read_input_token_cost": 5.5e-07, "cache_read_input_token_cost_above_272k_tokens": 1.1e-06, + "cache_read_input_token_cost_above_272k_tokens_priority": 2.2e-06, "cache_read_input_token_cost_priority": 1.1e-06, "deprecation_date": "2028-01-11", "input_cost_per_token": 5.5e-06, "input_cost_per_token_above_272k_tokens": 1.1e-05, + "input_cost_per_token_above_272k_tokens_priority": 2.2e-05, "input_cost_per_token_priority": 1.1e-05, "litellm_provider": "azure", "max_input_tokens": 922000, @@ -7081,6 +7104,7 @@ "mode": "chat", "output_cost_per_token": 3.3e-05, "output_cost_per_token_above_272k_tokens": 4.95e-05, + "output_cost_per_token_above_272k_tokens_priority": 9.9e-05, "output_cost_per_token_priority": 6.6e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, @@ -7117,13 +7141,16 @@ "azure/eu/gpt-5.6-terra": { "cache_creation_input_token_cost": 2.75e-06, "cache_creation_input_token_cost_above_272k_tokens": 5.5e-06, + "cache_creation_input_token_cost_above_272k_tokens_priority": 1.1e-05, "cache_creation_input_token_cost_priority": 5.5e-06, "cache_read_input_token_cost": 2.2e-07, "cache_read_input_token_cost_above_272k_tokens": 4.4e-07, + "cache_read_input_token_cost_above_272k_tokens_priority": 8.8e-07, "cache_read_input_token_cost_priority": 4.4e-07, "deprecation_date": "2028-01-11", "input_cost_per_token": 2.2e-06, "input_cost_per_token_above_272k_tokens": 4.4e-06, + "input_cost_per_token_above_272k_tokens_priority": 8.8e-06, "input_cost_per_token_priority": 4.4e-06, "litellm_provider": "azure", "max_input_tokens": 922000, @@ -7132,6 +7159,7 @@ "mode": "chat", "output_cost_per_token": 1.32e-05, "output_cost_per_token_above_272k_tokens": 1.98e-05, + "output_cost_per_token_above_272k_tokens_priority": 3.96e-05, "output_cost_per_token_priority": 2.64e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, @@ -7168,13 +7196,16 @@ "azure/eu/gpt-5.6-luna": { "cache_creation_input_token_cost": 2.75e-07, "cache_creation_input_token_cost_above_272k_tokens": 5.5e-07, + "cache_creation_input_token_cost_above_272k_tokens_priority": 1.1e-06, "cache_creation_input_token_cost_priority": 5.5e-07, "cache_read_input_token_cost": 2.2e-08, "cache_read_input_token_cost_above_272k_tokens": 4.4e-08, + "cache_read_input_token_cost_above_272k_tokens_priority": 8.8e-08, "cache_read_input_token_cost_priority": 4.4e-08, "deprecation_date": "2028-01-11", "input_cost_per_token": 2.2e-07, "input_cost_per_token_above_272k_tokens": 4.4e-07, + "input_cost_per_token_above_272k_tokens_priority": 8.8e-07, "input_cost_per_token_priority": 4.4e-07, "litellm_provider": "azure", "max_input_tokens": 922000, @@ -7183,6 +7214,7 @@ "mode": "chat", "output_cost_per_token": 1.32e-06, "output_cost_per_token_above_272k_tokens": 1.98e-06, + "output_cost_per_token_above_272k_tokens_priority": 3.96e-06, "output_cost_per_token_priority": 2.64e-06, "search_context_cost_per_query": { "search_context_size_high": 0.01, diff --git a/model_prices_and_context_window.json b/model_prices_and_context_window.json index 3f1d84cae97..57facc96319 100644 --- a/model_prices_and_context_window.json +++ b/model_prices_and_context_window.json @@ -6813,12 +6813,15 @@ "azure/us/gpt-5.6": { "cache_creation_input_token_cost": 6.875e-06, "cache_creation_input_token_cost_above_272k_tokens": 1.375e-05, + "cache_creation_input_token_cost_above_272k_tokens_priority": 2.75e-05, "cache_creation_input_token_cost_priority": 1.375e-05, "cache_read_input_token_cost": 5.5e-07, "cache_read_input_token_cost_above_272k_tokens": 1.1e-06, + "cache_read_input_token_cost_above_272k_tokens_priority": 2.2e-06, "cache_read_input_token_cost_priority": 1.1e-06, "input_cost_per_token": 5.5e-06, "input_cost_per_token_above_272k_tokens": 1.1e-05, + "input_cost_per_token_above_272k_tokens_priority": 2.2e-05, "input_cost_per_token_priority": 1.1e-05, "litellm_provider": "azure", "max_input_tokens": 922000, @@ -6827,6 +6830,7 @@ "mode": "chat", "output_cost_per_token": 3.3e-05, "output_cost_per_token_above_272k_tokens": 4.95e-05, + "output_cost_per_token_above_272k_tokens_priority": 9.9e-05, "output_cost_per_token_priority": 6.6e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, @@ -6863,13 +6867,16 @@ "azure/us/gpt-5.6-sol": { "cache_creation_input_token_cost": 6.875e-06, "cache_creation_input_token_cost_above_272k_tokens": 1.375e-05, + "cache_creation_input_token_cost_above_272k_tokens_priority": 2.75e-05, "cache_creation_input_token_cost_priority": 1.375e-05, "cache_read_input_token_cost": 5.5e-07, "cache_read_input_token_cost_above_272k_tokens": 1.1e-06, + "cache_read_input_token_cost_above_272k_tokens_priority": 2.2e-06, "cache_read_input_token_cost_priority": 1.1e-06, "deprecation_date": "2028-01-11", "input_cost_per_token": 5.5e-06, "input_cost_per_token_above_272k_tokens": 1.1e-05, + "input_cost_per_token_above_272k_tokens_priority": 2.2e-05, "input_cost_per_token_priority": 1.1e-05, "litellm_provider": "azure", "max_input_tokens": 922000, @@ -6878,6 +6885,7 @@ "mode": "chat", "output_cost_per_token": 3.3e-05, "output_cost_per_token_above_272k_tokens": 4.95e-05, + "output_cost_per_token_above_272k_tokens_priority": 9.9e-05, "output_cost_per_token_priority": 6.6e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, @@ -6914,13 +6922,16 @@ "azure/us/gpt-5.6-terra": { "cache_creation_input_token_cost": 2.75e-06, "cache_creation_input_token_cost_above_272k_tokens": 5.5e-06, + "cache_creation_input_token_cost_above_272k_tokens_priority": 1.1e-05, "cache_creation_input_token_cost_priority": 5.5e-06, "cache_read_input_token_cost": 2.2e-07, "cache_read_input_token_cost_above_272k_tokens": 4.4e-07, + "cache_read_input_token_cost_above_272k_tokens_priority": 8.8e-07, "cache_read_input_token_cost_priority": 4.4e-07, "deprecation_date": "2028-01-11", "input_cost_per_token": 2.2e-06, "input_cost_per_token_above_272k_tokens": 4.4e-06, + "input_cost_per_token_above_272k_tokens_priority": 8.8e-06, "input_cost_per_token_priority": 4.4e-06, "litellm_provider": "azure", "max_input_tokens": 922000, @@ -6929,6 +6940,7 @@ "mode": "chat", "output_cost_per_token": 1.32e-05, "output_cost_per_token_above_272k_tokens": 1.98e-05, + "output_cost_per_token_above_272k_tokens_priority": 3.96e-05, "output_cost_per_token_priority": 2.64e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, @@ -6965,13 +6977,16 @@ "azure/us/gpt-5.6-luna": { "cache_creation_input_token_cost": 2.75e-07, "cache_creation_input_token_cost_above_272k_tokens": 5.5e-07, + "cache_creation_input_token_cost_above_272k_tokens_priority": 1.1e-06, "cache_creation_input_token_cost_priority": 5.5e-07, "cache_read_input_token_cost": 2.2e-08, "cache_read_input_token_cost_above_272k_tokens": 4.4e-08, + "cache_read_input_token_cost_above_272k_tokens_priority": 8.8e-08, "cache_read_input_token_cost_priority": 4.4e-08, "deprecation_date": "2028-01-11", "input_cost_per_token": 2.2e-07, "input_cost_per_token_above_272k_tokens": 4.4e-07, + "input_cost_per_token_above_272k_tokens_priority": 8.8e-07, "input_cost_per_token_priority": 4.4e-07, "litellm_provider": "azure", "max_input_tokens": 922000, @@ -6980,6 +6995,7 @@ "mode": "chat", "output_cost_per_token": 1.32e-06, "output_cost_per_token_above_272k_tokens": 1.98e-06, + "output_cost_per_token_above_272k_tokens_priority": 3.96e-06, "output_cost_per_token_priority": 2.64e-06, "search_context_cost_per_query": { "search_context_size_high": 0.01, @@ -7016,12 +7032,15 @@ "azure/eu/gpt-5.6": { "cache_creation_input_token_cost": 6.875e-06, "cache_creation_input_token_cost_above_272k_tokens": 1.375e-05, + "cache_creation_input_token_cost_above_272k_tokens_priority": 2.75e-05, "cache_creation_input_token_cost_priority": 1.375e-05, "cache_read_input_token_cost": 5.5e-07, "cache_read_input_token_cost_above_272k_tokens": 1.1e-06, + "cache_read_input_token_cost_above_272k_tokens_priority": 2.2e-06, "cache_read_input_token_cost_priority": 1.1e-06, "input_cost_per_token": 5.5e-06, "input_cost_per_token_above_272k_tokens": 1.1e-05, + "input_cost_per_token_above_272k_tokens_priority": 2.2e-05, "input_cost_per_token_priority": 1.1e-05, "litellm_provider": "azure", "max_input_tokens": 922000, @@ -7030,6 +7049,7 @@ "mode": "chat", "output_cost_per_token": 3.3e-05, "output_cost_per_token_above_272k_tokens": 4.95e-05, + "output_cost_per_token_above_272k_tokens_priority": 9.9e-05, "output_cost_per_token_priority": 6.6e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, @@ -7066,13 +7086,16 @@ "azure/eu/gpt-5.6-sol": { "cache_creation_input_token_cost": 6.875e-06, "cache_creation_input_token_cost_above_272k_tokens": 1.375e-05, + "cache_creation_input_token_cost_above_272k_tokens_priority": 2.75e-05, "cache_creation_input_token_cost_priority": 1.375e-05, "cache_read_input_token_cost": 5.5e-07, "cache_read_input_token_cost_above_272k_tokens": 1.1e-06, + "cache_read_input_token_cost_above_272k_tokens_priority": 2.2e-06, "cache_read_input_token_cost_priority": 1.1e-06, "deprecation_date": "2028-01-11", "input_cost_per_token": 5.5e-06, "input_cost_per_token_above_272k_tokens": 1.1e-05, + "input_cost_per_token_above_272k_tokens_priority": 2.2e-05, "input_cost_per_token_priority": 1.1e-05, "litellm_provider": "azure", "max_input_tokens": 922000, @@ -7081,6 +7104,7 @@ "mode": "chat", "output_cost_per_token": 3.3e-05, "output_cost_per_token_above_272k_tokens": 4.95e-05, + "output_cost_per_token_above_272k_tokens_priority": 9.9e-05, "output_cost_per_token_priority": 6.6e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, @@ -7117,13 +7141,16 @@ "azure/eu/gpt-5.6-terra": { "cache_creation_input_token_cost": 2.75e-06, "cache_creation_input_token_cost_above_272k_tokens": 5.5e-06, + "cache_creation_input_token_cost_above_272k_tokens_priority": 1.1e-05, "cache_creation_input_token_cost_priority": 5.5e-06, "cache_read_input_token_cost": 2.2e-07, "cache_read_input_token_cost_above_272k_tokens": 4.4e-07, + "cache_read_input_token_cost_above_272k_tokens_priority": 8.8e-07, "cache_read_input_token_cost_priority": 4.4e-07, "deprecation_date": "2028-01-11", "input_cost_per_token": 2.2e-06, "input_cost_per_token_above_272k_tokens": 4.4e-06, + "input_cost_per_token_above_272k_tokens_priority": 8.8e-06, "input_cost_per_token_priority": 4.4e-06, "litellm_provider": "azure", "max_input_tokens": 922000, @@ -7132,6 +7159,7 @@ "mode": "chat", "output_cost_per_token": 1.32e-05, "output_cost_per_token_above_272k_tokens": 1.98e-05, + "output_cost_per_token_above_272k_tokens_priority": 3.96e-05, "output_cost_per_token_priority": 2.64e-05, "search_context_cost_per_query": { "search_context_size_high": 0.01, @@ -7168,13 +7196,16 @@ "azure/eu/gpt-5.6-luna": { "cache_creation_input_token_cost": 2.75e-07, "cache_creation_input_token_cost_above_272k_tokens": 5.5e-07, + "cache_creation_input_token_cost_above_272k_tokens_priority": 1.1e-06, "cache_creation_input_token_cost_priority": 5.5e-07, "cache_read_input_token_cost": 2.2e-08, "cache_read_input_token_cost_above_272k_tokens": 4.4e-08, + "cache_read_input_token_cost_above_272k_tokens_priority": 8.8e-08, "cache_read_input_token_cost_priority": 4.4e-08, "deprecation_date": "2028-01-11", "input_cost_per_token": 2.2e-07, "input_cost_per_token_above_272k_tokens": 4.4e-07, + "input_cost_per_token_above_272k_tokens_priority": 8.8e-07, "input_cost_per_token_priority": 4.4e-07, "litellm_provider": "azure", "max_input_tokens": 922000, @@ -7183,6 +7214,7 @@ "mode": "chat", "output_cost_per_token": 1.32e-06, "output_cost_per_token_above_272k_tokens": 1.98e-06, + "output_cost_per_token_above_272k_tokens_priority": 3.96e-06, "output_cost_per_token_priority": 2.64e-06, "search_context_cost_per_query": { "search_context_size_high": 0.01, diff --git a/tests/test_litellm/test_cost_calculator.py b/tests/test_litellm/test_cost_calculator.py index cd58c780710..5c731d63cda 100644 --- a/tests/test_litellm/test_cost_calculator.py +++ b/tests/test_litellm/test_cost_calculator.py @@ -1777,12 +1777,11 @@ def test_azure_gpt_5_6_rates_match_azure_price_page(_local_model_cost_map, model zone = model.split("/")[1] if zone in ("us", "eu"): global_entry = litellm.model_cost["azure/" + model.split("/", 2)[2]] - token_cost_keys = [ - key - for key in entry - if key.startswith(("input_cost_per_token", "output_cost_per_token", "cache_read", "cache_creation")) - ] + prefixes = ("input_cost_per_token", "output_cost_per_token", "cache_read", "cache_creation") + token_cost_keys = [key for key in entry if key.startswith(prefixes)] + global_token_cost_keys = [key for key in global_entry if key.startswith(prefixes)] assert len(token_cost_keys) >= 9 + assert sorted(token_cost_keys) == sorted(global_token_cost_keys) for key in token_cost_keys: assert entry[key] == pytest.approx(global_entry[key] * 1.1), key From c5e3b21019db0727431f0cf95516c10e417ae2e8 Mon Sep 17 00:00:00 2001 From: mateo-berri <277851410+mateo-berri@users.noreply.github.com> Date: Wed, 26 Aug 2026 11:51:57 -0700 Subject: [PATCH 4/4] test(utils): add priority cache write tier key to intended map schema --- tests/test_litellm/test_utils.py | 3 +++ 1 file changed, 3 insertions(+) diff --git a/tests/test_litellm/test_utils.py b/tests/test_litellm/test_utils.py index e1ea6edb2d8..f0d9e62f6f8 100644 --- a/tests/test_litellm/test_utils.py +++ b/tests/test_litellm/test_utils.py @@ -845,6 +845,9 @@ def test_aaamodel_prices_and_context_window_json_is_valid(): "cache_creation_input_token_cost_above_272k_tokens_flex": { "type": "number" }, + "cache_creation_input_token_cost_above_272k_tokens_priority": { + "type": "number" + }, "cache_creation_input_token_cost_flex": {"type": "number"}, "cache_creation_input_token_cost_priority": {"type": "number"}, "cache_read_input_token_cost": {"type": "number"},