From 5215c2d3f77cb7cad6db861f323a4e892f5c16bf Mon Sep 17 00:00:00 2001 From: Kerry Lu Date: Wed, 28 Oct 2026 14:59:23 -0700 Subject: [PATCH] fix(pricing): add databricks priority pay per token prices (#48891) * fix(pricing): add databricks priority pay per token prices * fix(pricing): bill gemini 3.6 flash cache writes at its databricks input rate * fix(pricing): drop image token priority rate the cost map schema does not carry --------- Co-authored-by: devin-ai-integration[bot] <158243242+devin-ai-integration[bot]@users.noreply.github.com> --- ...odel_prices_and_context_window_backup.json | 233 +++++++++++++++--- model_prices_and_context_window.json | 233 +++++++++++++++--- .../test_databricks_cost_calculator.py | 26 ++ 3 files changed, 428 insertions(+), 64 deletions(-) diff --git a/litellm/model_prices_and_context_window_backup.json b/litellm/model_prices_and_context_window_backup.json index 3e0e86012a4..1f2bd7d7837 100644 --- a/litellm/model_prices_and_context_window_backup.json +++ b/litellm/model_prices_and_context_window_backup.json @@ -21589,7 +21589,13 @@ ], "supports_vision": true, "supports_audio_input": true, - "supports_video_input": true + "supports_video_input": true, + "input_cost_per_token_priority": 5.6252e-07, + "output_cost_per_token_priority": 3.37498e-06, + "cache_read_input_token_cost_priority": 5.628e-08, + "cache_creation_input_token_cost_priority": 5.6252e-07, + "input_cost_per_audio_token_priority": 1.12497e-06, + "cache_read_input_audio_token_cost_priority": 1.1249e-07 }, "databricks/databricks-gemini-3-1-flash-image": { "cache_creation_input_token_cost": 6.2503e-07, @@ -21676,7 +21682,14 @@ "supports_reasoning": true, "supports_tool_choice": true, "supports_video_input": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 4.50002e-06, + "output_cost_per_token_priority": 2.699998e-05, + "cache_read_input_token_cost_priority": 4.5003e-07, + "cache_creation_input_token_cost_priority": 4.50002e-06, + "input_cost_per_token_above_200k_tokens_priority": 8.99997e-06, + "output_cost_per_token_above_200k_tokens_priority": 4.049997e-05, + "cache_read_input_token_cost_above_200k_tokens_priority": 8.9999e-07 }, "databricks/databricks-gemini-3-flash": { "cache_creation_input_token_cost": 6.2503e-07, @@ -21710,7 +21723,13 @@ ], "supports_vision": true, "supports_audio_input": true, - "supports_video_input": true + "supports_video_input": true, + "input_cost_per_token_priority": 1.12497e-06, + "output_cost_per_token_priority": 6.75003e-06, + "cache_read_input_token_cost_priority": 1.1249e-07, + "cache_creation_input_token_cost_priority": 1.12497e-06, + "input_cost_per_audio_token_priority": 2.25001e-06, + "cache_read_input_audio_token_cost_priority": 2.2498e-07 }, "databricks/databricks-gemini-3-pro": { "cache_creation_input_token_cost": 2.49998e-06, @@ -21757,7 +21776,11 @@ "supports_reasoning": true, "supports_sampling_params": false, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 1.35002e-06, + "output_cost_per_token_priority": 6.75003e-06, + "cache_read_input_token_cost_priority": 1.3503e-07, + "cache_creation_input_token_cost_priority": 1.35002e-06 }, "databricks/databricks-gemini-3-7-flash": { "cache_creation_input_token_cost": 7.4998e-07, @@ -21785,10 +21808,14 @@ "supports_reasoning": true, "supports_sampling_params": false, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 1.35002e-06, + "output_cost_per_token_priority": 6.75003e-06, + "cache_read_input_token_cost_priority": 1.3503e-07, + "cache_creation_input_token_cost_priority": 1.35002e-06 }, "databricks/databricks-gemini-3-6-flash": { - "cache_creation_input_token_cost": 1.87502e-06, + "cache_creation_input_token_cost": 7.4998e-07, "cache_read_input_token_cost": 7.497e-08, "input_cost_per_token": 7.4998e-07, "input_dbu_cost_per_token": 1.0714e-05, @@ -21816,7 +21843,11 @@ "supports_reasoning": true, "supports_sampling_params": false, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 1.35002e-06, + "output_cost_per_token_priority": 6.75003e-06, + "cache_read_input_token_cost_priority": 1.3503e-07, + "cache_creation_input_token_cost_priority": 1.35002e-06 }, "databricks/databricks-gemini-3-5-flash": { "cache_creation_input_token_cost": 1.87502e-06, @@ -21846,7 +21877,11 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 3.37498e-06, + "output_cost_per_token_priority": 2.025002e-05, + "cache_read_input_token_cost_priority": 3.3747e-07, + "cache_creation_input_token_cost_priority": 3.37498e-06 }, "databricks/databricks-gemini-3-5-flash-lite": { "cache_creation_input_token_cost": 3.7499e-07, @@ -21878,7 +21913,11 @@ "supports_response_schema": true, "supports_sampling_params": false, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 6.7501e-07, + "output_cost_per_token_priority": 5.62499e-06, + "cache_read_input_token_cost_priority": 6.748e-08, + "cache_creation_input_token_cost_priority": 6.7501e-07 }, "databricks/databricks-gemma-3-12b": { "cache_creation_input_token_cost": 1.5001e-07, @@ -21921,7 +21960,11 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": false + "supports_vision": false, + "input_cost_per_token_priority": 2.45e-06, + "output_cost_per_token_priority": 7.7e-06, + "cache_read_input_token_cost_priority": 4.55e-07, + "cache_creation_input_token_cost_priority": 2.45e-06 }, "databricks/databricks-glm-5-3": { "cache_creation_input_token_cost": 1.4e-06, @@ -21951,7 +21994,11 @@ "supports_response_schema": true, "supports_tool_choice": true, "supports_vision": false, - "thinking_always_on": true + "thinking_always_on": true, + "input_cost_per_token_priority": 2.45e-06, + "output_cost_per_token_priority": 7.7e-06, + "cache_read_input_token_cost_priority": 4.55e-07, + "cache_creation_input_token_cost_priority": 2.45e-06 }, "databricks/databricks-glm-5-3-flash": { "cache_creation_input_token_cost": 1.5001e-07, @@ -21980,7 +22027,11 @@ "supports_reasoning": true, "supports_tool_choice": true, "supports_vision": true, - "thinking_always_on": true + "thinking_always_on": true, + "input_cost_per_token_priority": 2.625e-07, + "output_cost_per_token_priority": 8.75e-07, + "cache_read_input_token_cost_priority": 5.25e-08, + "cache_creation_input_token_cost_priority": 2.625e-07 }, "databricks/databricks-gpt-5": { "cache_creation_input_token_cost": 1.24999e-06, @@ -21999,7 +22050,11 @@ "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", "supports_function_calling": true, "supports_prompt_caching": true, - "supports_tool_choice": true + "supports_tool_choice": true, + "input_cost_per_token_priority": 2.49998e-06, + "output_cost_per_token_priority": 1.999998e-05, + "cache_read_input_token_cost_priority": 2.4997e-07, + "cache_creation_input_token_cost_priority": 2.49998e-06 }, "databricks/databricks-gpt-5-1": { "cache_creation_input_token_cost": 1.24999e-06, @@ -22018,7 +22073,11 @@ "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", "supports_function_calling": true, "supports_prompt_caching": true, - "supports_tool_choice": true + "supports_tool_choice": true, + "input_cost_per_token_priority": 2.49998e-06, + "output_cost_per_token_priority": 1.999998e-05, + "cache_read_input_token_cost_priority": 2.4997e-07, + "cache_creation_input_token_cost_priority": 2.49998e-06 }, "databricks/databricks-gpt-5-2": { "cache_creation_input_token_cost": 1.75e-06, @@ -22037,7 +22096,11 @@ "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", "supports_function_calling": true, "supports_prompt_caching": true, - "supports_tool_choice": true + "supports_tool_choice": true, + "input_cost_per_token_priority": 3.5e-06, + "output_cost_per_token_priority": 2.8e-05, + "cache_read_input_token_cost_priority": 3.5e-07, + "cache_creation_input_token_cost_priority": 3.5e-06 }, "databricks/databricks-gpt-5-3-codex": { "cache_creation_input_token_cost": 1.75e-06, @@ -22056,7 +22119,11 @@ "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", "supports_function_calling": true, "supports_prompt_caching": true, - "supports_tool_choice": true + "supports_tool_choice": true, + "input_cost_per_token_priority": 3.5e-06, + "output_cost_per_token_priority": 2.8e-05, + "cache_read_input_token_cost_priority": 3.5e-07, + "cache_creation_input_token_cost_priority": 3.5e-06 }, "databricks/databricks-gpt-5-4": { "cache_creation_input_token_cost": 2.49998e-06, @@ -22087,7 +22154,14 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 5.00003e-06, + "output_cost_per_token_priority": 2.999997e-05, + "cache_read_input_token_cost_priority": 5.0001e-07, + "cache_creation_input_token_cost_priority": 5.00003e-06, + "input_cost_per_token_above_272k_tokens_priority": 5.00003e-06, + "output_cost_per_token_above_272k_tokens_priority": 2.999997e-05, + "cache_read_input_token_cost_above_272k_tokens_priority": 5.0001e-07 }, "databricks/databricks-gpt-5-4-mini": { "cache_creation_input_token_cost": 7.4998e-07, @@ -22115,7 +22189,11 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 1.50003e-06, + "output_cost_per_token_priority": 8.99997e-06, + "cache_read_input_token_cost_priority": 1.5001e-07, + "cache_creation_input_token_cost_priority": 1.50003e-06 }, "databricks/databricks-gpt-5-4-nano": { "cache_creation_input_token_cost": 1.9999e-07, @@ -22175,7 +22253,15 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 8.00002e-06, + "output_cost_per_token_priority": 4.000003e-05, + "cache_read_input_token_cost_priority": 8.0003e-07, + "cache_creation_input_token_cost_priority": 9.99999e-06, + "input_cost_per_token_above_272k_tokens_priority": 1.599997e-05, + "output_cost_per_token_above_272k_tokens_priority": 6.000001e-05, + "cache_read_input_token_cost_above_272k_tokens_priority": 1.59999e-06, + "cache_creation_input_token_cost_above_272k_tokens_priority": 1.999998e-05 }, "databricks/databricks-gpt-5-6-terra": { "cache_creation_input_token_cost": 2.49998e-06, @@ -22207,7 +22293,15 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 4.00001e-06, + "output_cost_per_token_priority": 2.399999e-05, + "cache_read_input_token_cost_priority": 3.9998e-07, + "cache_creation_input_token_cost_priority": 5.00003e-06, + "input_cost_per_token_above_272k_tokens_priority": 8.00002e-06, + "output_cost_per_token_above_272k_tokens_priority": 3.600002e-05, + "cache_read_input_token_cost_above_272k_tokens_priority": 8.0003e-07, + "cache_creation_input_token_cost_above_272k_tokens_priority": 9.99999e-06 }, "databricks/databricks-gpt-5-6-luna": { "cache_creation_input_token_cost": 2.4997e-07, @@ -22239,7 +22333,15 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 3.9998e-07, + "output_cost_per_token_priority": 2.40002e-06, + "cache_read_input_token_cost_priority": 3.997e-08, + "cache_creation_input_token_cost_priority": 5.0001e-07, + "input_cost_per_token_above_272k_tokens_priority": 8.0003e-07, + "output_cost_per_token_above_272k_tokens_priority": 3.60003e-06, + "cache_read_input_token_cost_above_272k_tokens_priority": 8.001e-08, + "cache_creation_input_token_cost_above_272k_tokens_priority": 1.00002e-06 }, "databricks/databricks-gpt-5-5": { "cache_creation_input_token_cost": 5.00003e-06, @@ -22270,7 +22372,14 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 1.249997e-05, + "output_cost_per_token_priority": 7.50001e-05, + "cache_read_input_token_cost_priority": 1.24999e-06, + "cache_creation_input_token_cost_priority": 1.249997e-05, + "input_cost_per_token_above_272k_tokens_priority": 1.249997e-05, + "output_cost_per_token_above_272k_tokens_priority": 7.50001e-05, + "cache_read_input_token_cost_above_272k_tokens_priority": 1.24999e-06 }, "databricks/databricks-gpt-5-5-pro": { "cache_creation_input_token_cost": 2.999997e-05, @@ -22319,7 +22428,11 @@ "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", "supports_function_calling": true, "supports_prompt_caching": true, - "supports_tool_choice": true + "supports_tool_choice": true, + "input_cost_per_token_priority": 4.5003e-07, + "output_cost_per_token_priority": 3.60003e-06, + "cache_read_input_token_cost_priority": 4.501e-08, + "cache_creation_input_token_cost_priority": 4.5003e-07 }, "databricks/databricks-gpt-5-nano": { "cache_creation_input_token_cost": 4.998e-08, @@ -22469,7 +22582,11 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 5.25e-06, + "output_cost_per_token_priority": 2.625e-05, + "cache_read_input_token_cost_priority": 5.25e-07, + "cache_creation_input_token_cost_priority": 5.25e-06 }, "databricks/system.ai.kimi-k3": { "litellm_provider": "databricks", @@ -22551,7 +22668,11 @@ "supports_reasoning": true, "supports_tool_choice": true, "supports_vision": false, - "thinking_always_on": true + "thinking_always_on": true, + "input_cost_per_token_priority": 3.85e-07, + "output_cost_per_token_priority": 3.85e-06, + "cache_read_input_token_cost_priority": 3.85e-07, + "cache_creation_input_token_cost_priority": 3.85e-07 }, "databricks/databricks-qwen3-next-80b-a3b-instruct": { "cache_creation_input_token_cost": 1.5001e-07, @@ -77774,7 +77895,11 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 5.25e-07, + "output_cost_per_token_priority": 2.1e-06, + "cache_read_input_token_cost_priority": 5.25e-08, + "cache_creation_input_token_cost_priority": 5.25e-07 }, "databricks/databricks-deepseek-v4-1-flash-1017": { "cache_creation_input_token_cost": 3.0002e-07, @@ -77897,7 +78022,11 @@ "supports_reasoning": true, "supports_tool_choice": true, "supports_video_input": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 4.50002e-06, + "output_cost_per_token_priority": 2.250003e-05, + "cache_read_input_token_cost_priority": 2.2498e-07, + "cache_creation_input_token_cost_priority": 4.50002e-06 }, "databricks/databricks-gpt-5-4-pro": { "cache_creation_input_token_cost": 2.999997e-05, @@ -77959,7 +78088,15 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 4.00001e-06, + "output_cost_per_token_priority": 1.99997e-05, + "cache_read_input_token_cost_priority": 1.9999e-07, + "cache_creation_input_token_cost_priority": 5.00003e-06, + "input_cost_per_token_above_272k_tokens_priority": 8.00002e-06, + "output_cost_per_token_above_272k_tokens_priority": 2.99999e-05, + "cache_read_input_token_cost_above_272k_tokens_priority": 3.9998e-07, + "cache_creation_input_token_cost_above_272k_tokens_priority": 9.99999e-06 }, "databricks/databricks-gpt-6-1-terra": { "cache_creation_input_token_cost": 1.24999e-06, @@ -77991,7 +78128,15 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 1.99997e-06, + "output_cost_per_token_priority": 9.99999e-06, + "cache_read_input_token_cost_priority": 1.0003e-07, + "cache_creation_input_token_cost_priority": 2.49998e-06, + "input_cost_per_token_above_272k_tokens_priority": 4.00001e-06, + "output_cost_per_token_above_272k_tokens_priority": 1.500002e-05, + "cache_read_input_token_cost_above_272k_tokens_priority": 1.9999e-07, + "cache_creation_input_token_cost_above_272k_tokens_priority": 5.00003e-06 }, "databricks/databricks-gpt-6-astra": { "cache_creation_input_token_cost": 1.249997e-05, @@ -78023,7 +78168,15 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 1.999998e-05, + "output_cost_per_token_priority": 9.99999e-05, + "cache_read_input_token_cost_priority": 1.99997e-06, + "cache_creation_input_token_cost_priority": 2.500001e-05, + "input_cost_per_token_above_272k_tokens_priority": 4.000003e-05, + "output_cost_per_token_above_272k_tokens_priority": 0.0001500002, + "cache_read_input_token_cost_above_272k_tokens_priority": 4.00001e-06, + "cache_creation_input_token_cost_above_272k_tokens_priority": 5.000002e-05 }, "databricks/databricks-gpt-6-luna": { "cache_creation_input_token_cost": 1.2502e-07, @@ -78055,7 +78208,15 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 1.9999e-07, + "output_cost_per_token_priority": 1.0003e-06, + "cache_read_input_token_cost_priority": 2.002e-08, + "cache_creation_input_token_cost_priority": 2.4997e-07, + "input_cost_per_token_above_272k_tokens_priority": 3.9998e-07, + "output_cost_per_token_above_272k_tokens_priority": 1.5001e-06, + "cache_read_input_token_cost_above_272k_tokens_priority": 3.997e-08, + "cache_creation_input_token_cost_above_272k_tokens_priority": 5.0001e-07 }, "databricks/databricks-gpt-6-sol": { "cache_creation_input_token_cost": 2.49998e-06, @@ -78087,7 +78248,15 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 4.00001e-06, + "output_cost_per_token_priority": 1.99997e-05, + "cache_read_input_token_cost_priority": 3.9998e-07, + "cache_creation_input_token_cost_priority": 5.00003e-06, + "input_cost_per_token_above_272k_tokens_priority": 8.00002e-06, + "output_cost_per_token_above_272k_tokens_priority": 2.99999e-05, + "cache_read_input_token_cost_above_272k_tokens_priority": 8.0003e-07, + "cache_creation_input_token_cost_above_272k_tokens_priority": 9.99999e-06 }, "databricks/databricks-gpt-image-1": { "cache_creation_input_token_cost": 5.00003e-06, diff --git a/model_prices_and_context_window.json b/model_prices_and_context_window.json index 3e0e86012a4..1f2bd7d7837 100644 --- a/model_prices_and_context_window.json +++ b/model_prices_and_context_window.json @@ -21589,7 +21589,13 @@ ], "supports_vision": true, "supports_audio_input": true, - "supports_video_input": true + "supports_video_input": true, + "input_cost_per_token_priority": 5.6252e-07, + "output_cost_per_token_priority": 3.37498e-06, + "cache_read_input_token_cost_priority": 5.628e-08, + "cache_creation_input_token_cost_priority": 5.6252e-07, + "input_cost_per_audio_token_priority": 1.12497e-06, + "cache_read_input_audio_token_cost_priority": 1.1249e-07 }, "databricks/databricks-gemini-3-1-flash-image": { "cache_creation_input_token_cost": 6.2503e-07, @@ -21676,7 +21682,14 @@ "supports_reasoning": true, "supports_tool_choice": true, "supports_video_input": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 4.50002e-06, + "output_cost_per_token_priority": 2.699998e-05, + "cache_read_input_token_cost_priority": 4.5003e-07, + "cache_creation_input_token_cost_priority": 4.50002e-06, + "input_cost_per_token_above_200k_tokens_priority": 8.99997e-06, + "output_cost_per_token_above_200k_tokens_priority": 4.049997e-05, + "cache_read_input_token_cost_above_200k_tokens_priority": 8.9999e-07 }, "databricks/databricks-gemini-3-flash": { "cache_creation_input_token_cost": 6.2503e-07, @@ -21710,7 +21723,13 @@ ], "supports_vision": true, "supports_audio_input": true, - "supports_video_input": true + "supports_video_input": true, + "input_cost_per_token_priority": 1.12497e-06, + "output_cost_per_token_priority": 6.75003e-06, + "cache_read_input_token_cost_priority": 1.1249e-07, + "cache_creation_input_token_cost_priority": 1.12497e-06, + "input_cost_per_audio_token_priority": 2.25001e-06, + "cache_read_input_audio_token_cost_priority": 2.2498e-07 }, "databricks/databricks-gemini-3-pro": { "cache_creation_input_token_cost": 2.49998e-06, @@ -21757,7 +21776,11 @@ "supports_reasoning": true, "supports_sampling_params": false, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 1.35002e-06, + "output_cost_per_token_priority": 6.75003e-06, + "cache_read_input_token_cost_priority": 1.3503e-07, + "cache_creation_input_token_cost_priority": 1.35002e-06 }, "databricks/databricks-gemini-3-7-flash": { "cache_creation_input_token_cost": 7.4998e-07, @@ -21785,10 +21808,14 @@ "supports_reasoning": true, "supports_sampling_params": false, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 1.35002e-06, + "output_cost_per_token_priority": 6.75003e-06, + "cache_read_input_token_cost_priority": 1.3503e-07, + "cache_creation_input_token_cost_priority": 1.35002e-06 }, "databricks/databricks-gemini-3-6-flash": { - "cache_creation_input_token_cost": 1.87502e-06, + "cache_creation_input_token_cost": 7.4998e-07, "cache_read_input_token_cost": 7.497e-08, "input_cost_per_token": 7.4998e-07, "input_dbu_cost_per_token": 1.0714e-05, @@ -21816,7 +21843,11 @@ "supports_reasoning": true, "supports_sampling_params": false, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 1.35002e-06, + "output_cost_per_token_priority": 6.75003e-06, + "cache_read_input_token_cost_priority": 1.3503e-07, + "cache_creation_input_token_cost_priority": 1.35002e-06 }, "databricks/databricks-gemini-3-5-flash": { "cache_creation_input_token_cost": 1.87502e-06, @@ -21846,7 +21877,11 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 3.37498e-06, + "output_cost_per_token_priority": 2.025002e-05, + "cache_read_input_token_cost_priority": 3.3747e-07, + "cache_creation_input_token_cost_priority": 3.37498e-06 }, "databricks/databricks-gemini-3-5-flash-lite": { "cache_creation_input_token_cost": 3.7499e-07, @@ -21878,7 +21913,11 @@ "supports_response_schema": true, "supports_sampling_params": false, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 6.7501e-07, + "output_cost_per_token_priority": 5.62499e-06, + "cache_read_input_token_cost_priority": 6.748e-08, + "cache_creation_input_token_cost_priority": 6.7501e-07 }, "databricks/databricks-gemma-3-12b": { "cache_creation_input_token_cost": 1.5001e-07, @@ -21921,7 +21960,11 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": false + "supports_vision": false, + "input_cost_per_token_priority": 2.45e-06, + "output_cost_per_token_priority": 7.7e-06, + "cache_read_input_token_cost_priority": 4.55e-07, + "cache_creation_input_token_cost_priority": 2.45e-06 }, "databricks/databricks-glm-5-3": { "cache_creation_input_token_cost": 1.4e-06, @@ -21951,7 +21994,11 @@ "supports_response_schema": true, "supports_tool_choice": true, "supports_vision": false, - "thinking_always_on": true + "thinking_always_on": true, + "input_cost_per_token_priority": 2.45e-06, + "output_cost_per_token_priority": 7.7e-06, + "cache_read_input_token_cost_priority": 4.55e-07, + "cache_creation_input_token_cost_priority": 2.45e-06 }, "databricks/databricks-glm-5-3-flash": { "cache_creation_input_token_cost": 1.5001e-07, @@ -21980,7 +22027,11 @@ "supports_reasoning": true, "supports_tool_choice": true, "supports_vision": true, - "thinking_always_on": true + "thinking_always_on": true, + "input_cost_per_token_priority": 2.625e-07, + "output_cost_per_token_priority": 8.75e-07, + "cache_read_input_token_cost_priority": 5.25e-08, + "cache_creation_input_token_cost_priority": 2.625e-07 }, "databricks/databricks-gpt-5": { "cache_creation_input_token_cost": 1.24999e-06, @@ -21999,7 +22050,11 @@ "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", "supports_function_calling": true, "supports_prompt_caching": true, - "supports_tool_choice": true + "supports_tool_choice": true, + "input_cost_per_token_priority": 2.49998e-06, + "output_cost_per_token_priority": 1.999998e-05, + "cache_read_input_token_cost_priority": 2.4997e-07, + "cache_creation_input_token_cost_priority": 2.49998e-06 }, "databricks/databricks-gpt-5-1": { "cache_creation_input_token_cost": 1.24999e-06, @@ -22018,7 +22073,11 @@ "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", "supports_function_calling": true, "supports_prompt_caching": true, - "supports_tool_choice": true + "supports_tool_choice": true, + "input_cost_per_token_priority": 2.49998e-06, + "output_cost_per_token_priority": 1.999998e-05, + "cache_read_input_token_cost_priority": 2.4997e-07, + "cache_creation_input_token_cost_priority": 2.49998e-06 }, "databricks/databricks-gpt-5-2": { "cache_creation_input_token_cost": 1.75e-06, @@ -22037,7 +22096,11 @@ "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", "supports_function_calling": true, "supports_prompt_caching": true, - "supports_tool_choice": true + "supports_tool_choice": true, + "input_cost_per_token_priority": 3.5e-06, + "output_cost_per_token_priority": 2.8e-05, + "cache_read_input_token_cost_priority": 3.5e-07, + "cache_creation_input_token_cost_priority": 3.5e-06 }, "databricks/databricks-gpt-5-3-codex": { "cache_creation_input_token_cost": 1.75e-06, @@ -22056,7 +22119,11 @@ "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", "supports_function_calling": true, "supports_prompt_caching": true, - "supports_tool_choice": true + "supports_tool_choice": true, + "input_cost_per_token_priority": 3.5e-06, + "output_cost_per_token_priority": 2.8e-05, + "cache_read_input_token_cost_priority": 3.5e-07, + "cache_creation_input_token_cost_priority": 3.5e-06 }, "databricks/databricks-gpt-5-4": { "cache_creation_input_token_cost": 2.49998e-06, @@ -22087,7 +22154,14 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 5.00003e-06, + "output_cost_per_token_priority": 2.999997e-05, + "cache_read_input_token_cost_priority": 5.0001e-07, + "cache_creation_input_token_cost_priority": 5.00003e-06, + "input_cost_per_token_above_272k_tokens_priority": 5.00003e-06, + "output_cost_per_token_above_272k_tokens_priority": 2.999997e-05, + "cache_read_input_token_cost_above_272k_tokens_priority": 5.0001e-07 }, "databricks/databricks-gpt-5-4-mini": { "cache_creation_input_token_cost": 7.4998e-07, @@ -22115,7 +22189,11 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 1.50003e-06, + "output_cost_per_token_priority": 8.99997e-06, + "cache_read_input_token_cost_priority": 1.5001e-07, + "cache_creation_input_token_cost_priority": 1.50003e-06 }, "databricks/databricks-gpt-5-4-nano": { "cache_creation_input_token_cost": 1.9999e-07, @@ -22175,7 +22253,15 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 8.00002e-06, + "output_cost_per_token_priority": 4.000003e-05, + "cache_read_input_token_cost_priority": 8.0003e-07, + "cache_creation_input_token_cost_priority": 9.99999e-06, + "input_cost_per_token_above_272k_tokens_priority": 1.599997e-05, + "output_cost_per_token_above_272k_tokens_priority": 6.000001e-05, + "cache_read_input_token_cost_above_272k_tokens_priority": 1.59999e-06, + "cache_creation_input_token_cost_above_272k_tokens_priority": 1.999998e-05 }, "databricks/databricks-gpt-5-6-terra": { "cache_creation_input_token_cost": 2.49998e-06, @@ -22207,7 +22293,15 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 4.00001e-06, + "output_cost_per_token_priority": 2.399999e-05, + "cache_read_input_token_cost_priority": 3.9998e-07, + "cache_creation_input_token_cost_priority": 5.00003e-06, + "input_cost_per_token_above_272k_tokens_priority": 8.00002e-06, + "output_cost_per_token_above_272k_tokens_priority": 3.600002e-05, + "cache_read_input_token_cost_above_272k_tokens_priority": 8.0003e-07, + "cache_creation_input_token_cost_above_272k_tokens_priority": 9.99999e-06 }, "databricks/databricks-gpt-5-6-luna": { "cache_creation_input_token_cost": 2.4997e-07, @@ -22239,7 +22333,15 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 3.9998e-07, + "output_cost_per_token_priority": 2.40002e-06, + "cache_read_input_token_cost_priority": 3.997e-08, + "cache_creation_input_token_cost_priority": 5.0001e-07, + "input_cost_per_token_above_272k_tokens_priority": 8.0003e-07, + "output_cost_per_token_above_272k_tokens_priority": 3.60003e-06, + "cache_read_input_token_cost_above_272k_tokens_priority": 8.001e-08, + "cache_creation_input_token_cost_above_272k_tokens_priority": 1.00002e-06 }, "databricks/databricks-gpt-5-5": { "cache_creation_input_token_cost": 5.00003e-06, @@ -22270,7 +22372,14 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 1.249997e-05, + "output_cost_per_token_priority": 7.50001e-05, + "cache_read_input_token_cost_priority": 1.24999e-06, + "cache_creation_input_token_cost_priority": 1.249997e-05, + "input_cost_per_token_above_272k_tokens_priority": 1.249997e-05, + "output_cost_per_token_above_272k_tokens_priority": 7.50001e-05, + "cache_read_input_token_cost_above_272k_tokens_priority": 1.24999e-06 }, "databricks/databricks-gpt-5-5-pro": { "cache_creation_input_token_cost": 2.999997e-05, @@ -22319,7 +22428,11 @@ "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", "supports_function_calling": true, "supports_prompt_caching": true, - "supports_tool_choice": true + "supports_tool_choice": true, + "input_cost_per_token_priority": 4.5003e-07, + "output_cost_per_token_priority": 3.60003e-06, + "cache_read_input_token_cost_priority": 4.501e-08, + "cache_creation_input_token_cost_priority": 4.5003e-07 }, "databricks/databricks-gpt-5-nano": { "cache_creation_input_token_cost": 4.998e-08, @@ -22469,7 +22582,11 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 5.25e-06, + "output_cost_per_token_priority": 2.625e-05, + "cache_read_input_token_cost_priority": 5.25e-07, + "cache_creation_input_token_cost_priority": 5.25e-06 }, "databricks/system.ai.kimi-k3": { "litellm_provider": "databricks", @@ -22551,7 +22668,11 @@ "supports_reasoning": true, "supports_tool_choice": true, "supports_vision": false, - "thinking_always_on": true + "thinking_always_on": true, + "input_cost_per_token_priority": 3.85e-07, + "output_cost_per_token_priority": 3.85e-06, + "cache_read_input_token_cost_priority": 3.85e-07, + "cache_creation_input_token_cost_priority": 3.85e-07 }, "databricks/databricks-qwen3-next-80b-a3b-instruct": { "cache_creation_input_token_cost": 1.5001e-07, @@ -77774,7 +77895,11 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 5.25e-07, + "output_cost_per_token_priority": 2.1e-06, + "cache_read_input_token_cost_priority": 5.25e-08, + "cache_creation_input_token_cost_priority": 5.25e-07 }, "databricks/databricks-deepseek-v4-1-flash-1017": { "cache_creation_input_token_cost": 3.0002e-07, @@ -77897,7 +78022,11 @@ "supports_reasoning": true, "supports_tool_choice": true, "supports_video_input": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 4.50002e-06, + "output_cost_per_token_priority": 2.250003e-05, + "cache_read_input_token_cost_priority": 2.2498e-07, + "cache_creation_input_token_cost_priority": 4.50002e-06 }, "databricks/databricks-gpt-5-4-pro": { "cache_creation_input_token_cost": 2.999997e-05, @@ -77959,7 +78088,15 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 4.00001e-06, + "output_cost_per_token_priority": 1.99997e-05, + "cache_read_input_token_cost_priority": 1.9999e-07, + "cache_creation_input_token_cost_priority": 5.00003e-06, + "input_cost_per_token_above_272k_tokens_priority": 8.00002e-06, + "output_cost_per_token_above_272k_tokens_priority": 2.99999e-05, + "cache_read_input_token_cost_above_272k_tokens_priority": 3.9998e-07, + "cache_creation_input_token_cost_above_272k_tokens_priority": 9.99999e-06 }, "databricks/databricks-gpt-6-1-terra": { "cache_creation_input_token_cost": 1.24999e-06, @@ -77991,7 +78128,15 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 1.99997e-06, + "output_cost_per_token_priority": 9.99999e-06, + "cache_read_input_token_cost_priority": 1.0003e-07, + "cache_creation_input_token_cost_priority": 2.49998e-06, + "input_cost_per_token_above_272k_tokens_priority": 4.00001e-06, + "output_cost_per_token_above_272k_tokens_priority": 1.500002e-05, + "cache_read_input_token_cost_above_272k_tokens_priority": 1.9999e-07, + "cache_creation_input_token_cost_above_272k_tokens_priority": 5.00003e-06 }, "databricks/databricks-gpt-6-astra": { "cache_creation_input_token_cost": 1.249997e-05, @@ -78023,7 +78168,15 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 1.999998e-05, + "output_cost_per_token_priority": 9.99999e-05, + "cache_read_input_token_cost_priority": 1.99997e-06, + "cache_creation_input_token_cost_priority": 2.500001e-05, + "input_cost_per_token_above_272k_tokens_priority": 4.000003e-05, + "output_cost_per_token_above_272k_tokens_priority": 0.0001500002, + "cache_read_input_token_cost_above_272k_tokens_priority": 4.00001e-06, + "cache_creation_input_token_cost_above_272k_tokens_priority": 5.000002e-05 }, "databricks/databricks-gpt-6-luna": { "cache_creation_input_token_cost": 1.2502e-07, @@ -78055,7 +78208,15 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 1.9999e-07, + "output_cost_per_token_priority": 1.0003e-06, + "cache_read_input_token_cost_priority": 2.002e-08, + "cache_creation_input_token_cost_priority": 2.4997e-07, + "input_cost_per_token_above_272k_tokens_priority": 3.9998e-07, + "output_cost_per_token_above_272k_tokens_priority": 1.5001e-06, + "cache_read_input_token_cost_above_272k_tokens_priority": 3.997e-08, + "cache_creation_input_token_cost_above_272k_tokens_priority": 5.0001e-07 }, "databricks/databricks-gpt-6-sol": { "cache_creation_input_token_cost": 2.49998e-06, @@ -78087,7 +78248,15 @@ "supports_prompt_caching": true, "supports_reasoning": true, "supports_tool_choice": true, - "supports_vision": true + "supports_vision": true, + "input_cost_per_token_priority": 4.00001e-06, + "output_cost_per_token_priority": 1.99997e-05, + "cache_read_input_token_cost_priority": 3.9998e-07, + "cache_creation_input_token_cost_priority": 5.00003e-06, + "input_cost_per_token_above_272k_tokens_priority": 8.00002e-06, + "output_cost_per_token_above_272k_tokens_priority": 2.99999e-05, + "cache_read_input_token_cost_above_272k_tokens_priority": 8.0003e-07, + "cache_creation_input_token_cost_above_272k_tokens_priority": 9.99999e-06 }, "databricks/databricks-gpt-image-1": { "cache_creation_input_token_cost": 5.00003e-06, diff --git a/tests/unit/llms/databricks/test_databricks_cost_calculator.py b/tests/unit/llms/databricks/test_databricks_cost_calculator.py index 34d3806d6eb..c35aa6e4cf0 100644 --- a/tests/unit/llms/databricks/test_databricks_cost_calculator.py +++ b/tests/unit/llms/databricks/test_databricks_cost_calculator.py @@ -253,3 +253,29 @@ def test_cost_per_token_bills_the_served_priority_tier( prompt_cost, completion_cost = cost_per_token(model="databricks/dbrx-tiered-test", usage=usage) assert prompt_cost == pytest.approx(30 * 0.001) assert completion_cost == pytest.approx(40 * 0.002) + + +RATES_WITHOUT_A_PRIORITY_FIELD: Final = frozenset({"input_cost_per_image_token"}) + + +def _is_standard_rate_key(key: str) -> bool: + return ( + ("_cost_per_" in key or key.endswith("_token_cost")) + and "dbu" not in key + and key not in RATES_WITHOUT_A_PRIORITY_FIELD + and not key.endswith(("_priority", "_flex", "_batches")) + ) + + +def test_every_databricks_priority_row_prices_every_standard_rate_at_the_priority_tier() -> None: + cost_map: Final = json.loads(MAIN_PRICES.read_text()) + priority_rows: Final = { + key: row + for key, row in cost_map.items() + if key.startswith("databricks/") and "input_cost_per_token_priority" in row + } + + assert priority_rows + for key, row in priority_rows.items(): + missing = sorted(f for f in row if _is_standard_rate_key(f) and f"{f}_priority" not in row) + assert not missing, f"{key} bills these at the standard rate on priority requests: {missing}"