diff --git a/litellm/model_prices_and_context_window_backup.json b/litellm/model_prices_and_context_window_backup.json
index b6eec1c9574..810fd22a64f 100644
--- a/litellm/model_prices_and_context_window_backup.json
+++ b/litellm/model_prices_and_context_window_backup.json
@@ -99949,17 +99949,20 @@
         ]
     },
     "databricks/databricks-claude-fable-5-2": {
-        "cache_creation_input_token_cost": 8e-06,
-        "cache_read_input_token_cost": 8e-06,
-        "input_cost_per_token": 8e-06,
+        "cache_creation_input_token_cost": 9.99999e-06,
+        "cache_creation_input_token_cost_above_1hr": 1.599997e-05,
+        "cache_read_input_token_cost": 1.9999e-07,
+        "input_cost_per_token": 8.00002e-06,
+        "input_dbu_cost_per_token": 0.000114286,
         "litellm_provider": "databricks",
         "max_input_tokens": 1000000,
         "max_output_tokens": 128000,
-        "metadata": {"notes":"Claude Fable 5.2 on Databricks is available only through Unity Gateway with the model service name system.ai.claude-fable-5-2, and Databricks hosts it through Amazon Bedrock. The Databricks pricing page lists no row for it yet, so input and output are inherited from the catalog Bedrock row anthropic.claude-fable-5-2, no '*_dbu_cost_per_token' fields are published, and cache tokens bill at the input rate until Databricks publishes cache DBU rates."},
+        "metadata": {"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},
         "mode": "chat",
-        "output_cost_per_token": 4e-05,
+        "output_cost_per_token": 4.000003e-05,
+        "output_dbu_cost_per_token": 0.000571429,
         "prompt_cache_min_tokens": 512.0,
-        "source": "https://docs.databricks.com/aws/en/machine-learning/foundation-model-apis/supported-models",
+        "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving",
         "supports_adaptive_thinking": true,
         "supports_anthropic_thinking_payload": true,
         "supports_assistant_prefill": false,
@@ -99974,17 +99977,20 @@
         "thinking_always_on": true
     },
     "databricks/system.ai.claude-fable-5-2": {
-        "cache_creation_input_token_cost": 8e-06,
-        "cache_read_input_token_cost": 8e-06,
-        "input_cost_per_token": 8e-06,
+        "cache_creation_input_token_cost": 9.99999e-06,
+        "cache_creation_input_token_cost_above_1hr": 1.599997e-05,
+        "cache_read_input_token_cost": 1.9999e-07,
+        "input_cost_per_token": 8.00002e-06,
+        "input_dbu_cost_per_token": 0.000114286,
         "litellm_provider": "databricks",
         "max_input_tokens": 1000000,
         "max_output_tokens": 128000,
-        "metadata": {"notes":"Claude Fable 5.2 on Databricks is available only through Unity Gateway with the model service name system.ai.claude-fable-5-2, and Databricks hosts it through Amazon Bedrock. The Databricks pricing page lists no row for it yet, so input and output are inherited from the catalog Bedrock row anthropic.claude-fable-5-2, no '*_dbu_cost_per_token' fields are published, and cache tokens bill at the input rate until Databricks publishes cache DBU rates."},
+        "metadata": {"notes":"Claude Fable 5.2 on Databricks is served through Unity Gateway with the model service name system.ai.claude-fable-5-2, hosted through Amazon Bedrock. Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},
         "mode": "chat",
-        "output_cost_per_token": 4e-05,
+        "output_cost_per_token": 4.000003e-05,
+        "output_dbu_cost_per_token": 0.000571429,
         "prompt_cache_min_tokens": 512.0,
-        "source": "https://docs.databricks.com/aws/en/machine-learning/foundation-model-apis/supported-models",
+        "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving",
         "supports_adaptive_thinking": true,
         "supports_anthropic_thinking_payload": true,
         "supports_assistant_prefill": false,
diff --git a/model_prices_and_context_window.json b/model_prices_and_context_window.json
index b6eec1c9574..810fd22a64f 100644
--- a/model_prices_and_context_window.json
+++ b/model_prices_and_context_window.json
@@ -99949,17 +99949,20 @@
         ]
     },
     "databricks/databricks-claude-fable-5-2": {
-        "cache_creation_input_token_cost": 8e-06,
-        "cache_read_input_token_cost": 8e-06,
-        "input_cost_per_token": 8e-06,
+        "cache_creation_input_token_cost": 9.99999e-06,
+        "cache_creation_input_token_cost_above_1hr": 1.599997e-05,
+        "cache_read_input_token_cost": 1.9999e-07,
+        "input_cost_per_token": 8.00002e-06,
+        "input_dbu_cost_per_token": 0.000114286,
         "litellm_provider": "databricks",
         "max_input_tokens": 1000000,
         "max_output_tokens": 128000,
-        "metadata": {"notes":"Claude Fable 5.2 on Databricks is available only through Unity Gateway with the model service name system.ai.claude-fable-5-2, and Databricks hosts it through Amazon Bedrock. The Databricks pricing page lists no row for it yet, so input and output are inherited from the catalog Bedrock row anthropic.claude-fable-5-2, no '*_dbu_cost_per_token' fields are published, and cache tokens bill at the input rate until Databricks publishes cache DBU rates."},
+        "metadata": {"notes":"Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},
         "mode": "chat",
-        "output_cost_per_token": 4e-05,
+        "output_cost_per_token": 4.000003e-05,
+        "output_dbu_cost_per_token": 0.000571429,
         "prompt_cache_min_tokens": 512.0,
-        "source": "https://docs.databricks.com/aws/en/machine-learning/foundation-model-apis/supported-models",
+        "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving",
         "supports_adaptive_thinking": true,
         "supports_anthropic_thinking_payload": true,
         "supports_assistant_prefill": false,
@@ -99974,17 +99977,20 @@
         "thinking_always_on": true
     },
     "databricks/system.ai.claude-fable-5-2": {
-        "cache_creation_input_token_cost": 8e-06,
-        "cache_read_input_token_cost": 8e-06,
-        "input_cost_per_token": 8e-06,
+        "cache_creation_input_token_cost": 9.99999e-06,
+        "cache_creation_input_token_cost_above_1hr": 1.599997e-05,
+        "cache_read_input_token_cost": 1.9999e-07,
+        "input_cost_per_token": 8.00002e-06,
+        "input_dbu_cost_per_token": 0.000114286,
         "litellm_provider": "databricks",
         "max_input_tokens": 1000000,
         "max_output_tokens": 128000,
-        "metadata": {"notes":"Claude Fable 5.2 on Databricks is available only through Unity Gateway with the model service name system.ai.claude-fable-5-2, and Databricks hosts it through Amazon Bedrock. The Databricks pricing page lists no row for it yet, so input and output are inherited from the catalog Bedrock row anthropic.claude-fable-5-2, no '*_dbu_cost_per_token' fields are published, and cache tokens bill at the input rate until Databricks publishes cache DBU rates."},
+        "metadata": {"notes":"Claude Fable 5.2 on Databricks is served through Unity Gateway with the model service name system.ai.claude-fable-5-2, hosted through Amazon Bedrock. Costs per token are the published Global DBU rates times $0.070 per DBU. The '*_dbu_cost_per_token' fields are provided for reference; cost calculation reads the dollar '*_cost_per_token' fields."},
         "mode": "chat",
-        "output_cost_per_token": 4e-05,
+        "output_cost_per_token": 4.000003e-05,
+        "output_dbu_cost_per_token": 0.000571429,
         "prompt_cache_min_tokens": 512.0,
-        "source": "https://docs.databricks.com/aws/en/machine-learning/foundation-model-apis/supported-models",
+        "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving",
         "supports_adaptive_thinking": true,
         "supports_anthropic_thinking_payload": true,
         "supports_assistant_prefill": false,
diff --git a/tests/unit/llms/databricks/test_databricks_cost_calculator.py b/tests/unit/llms/databricks/test_databricks_cost_calculator.py
index e9bc75119aa..b968191383d 100644
--- a/tests/unit/llms/databricks/test_databricks_cost_calculator.py
+++ b/tests/unit/llms/databricks/test_databricks_cost_calculator.py
@@ -44,6 +44,8 @@ PUBLISHED_DBU_PER_MILLION: Final = {
     "databricks/databricks-claude-opus-5-5": ("57.143", "285.714", "71.429", "2.857"),
     "databricks/databricks-claude-fable-5-1": ("142.858", "714.286", "178.572", "3.572"),
     "databricks/databricks-claude-fable-5": ("142.858", "714.286", "178.572", "14.286"),
+    "databricks/databricks-claude-fable-5-2": ("114.286", "571.429", "142.857", "2.857"),
+    "databricks/system.ai.claude-fable-5-2": ("114.286", "571.429", "142.857", "2.857"),
     "databricks/databricks-claude-opus-5": ("71.429", "357.143", "89.286", "7.143"),
     "databricks/databricks-claude-opus-4-8": ("71.429", "357.143", "89.286", "7.143"),
     "databricks/databricks-claude-opus-4-7": ("71.429", "357.143", "89.286", "7.143"),