From af2f447698bed16210c8bc452714876455d51ebb Mon Sep 17 00:00:00 2001
From: mubashir1osmani <mubashir.osmani777@gmail.com>
Date: Mon, 1 Sep 2025 10:28:48 -0400
Subject: [PATCH] added prometheus metrics

---
 .../integrations/prometheus.py                | 20 +++++++------
 litellm/types/integrations/prometheus.py      | 28 +++++++++++++++++--
 2 files changed, 36 insertions(+), 12 deletions(-)

diff --git a/enterprise/litellm_enterprise/integrations/prometheus.py b/enterprise/litellm_enterprise/integrations/prometheus.py
index a2d781fa1c..1aa2359bc1 100644
--- a/enterprise/litellm_enterprise/integrations/prometheus.py
+++ b/enterprise/litellm_enterprise/integrations/prometheus.py
@@ -95,13 +95,14 @@ class PrometheusLogger(CustomLogger):
             self.litellm_llm_api_time_to_first_token_metric = self._histogram_factory(
                 "litellm_llm_api_time_to_first_token_metric",
                 "Time to first token for a models LLM API call",
-                labelnames=[
-                    "model",
-                    "hashed_api_key",
-                    "api_key_alias",
-                    "team",
-                    "team_alias",
-                ],
+                # labelnames=[
+                #     "model",
+                #     "hashed_api_key",
+                #     "api_key_alias",
+                #     "team",
+                #     "team_alias",
+                # ],
+                labelnames=self.get_labels_for_metric("litellm_llm_api_time_to_first_token_metric"),
                 buckets=LATENCY_BUCKETS,
             )
 
@@ -255,13 +256,14 @@ class PrometheusLogger(CustomLogger):
             self.litellm_deployment_state = self._gauge_factory(
                 "litellm_deployment_state",
                 "LLM Deployment Analytics - The state of the deployment: 0 = healthy, 1 = partial outage, 2 = complete outage",
-                labelnames=_logged_llm_labels,
+                labelnames=self.get_labels_for_metric("litellm_deployment_state")
             )
 
             self.litellm_deployment_cooled_down = self._counter_factory(
                 "litellm_deployment_cooled_down",
                 "LLM Deployment Analytics - Number of times a deployment has been cooled down by LiteLLM load balancing logic. exception_status is the status of the exception that caused the deployment to be cooled down",
-                labelnames=_logged_llm_labels + [EXCEPTION_STATUS],
+                # labelnames=_logged_llm_labels + [EXCEPTION_STATUS],
+                labelnames=self.get_labels_for_metric("litellm_deployment_cooled_down")
             )
 
             self.litellm_deployment_success_responses = self._counter_factory(
diff --git a/litellm/types/integrations/prometheus.py b/litellm/types/integrations/prometheus.py
index e0ee950d26..4ab0ccf502 100644
--- a/litellm/types/integrations/prometheus.py
+++ b/litellm/types/integrations/prometheus.py
@@ -154,6 +154,7 @@ class UserAPIKeyLabelNames(Enum):
 
 DEFINED_PROMETHEUS_METRICS = Literal[
     "litellm_llm_api_latency_metric",
+    "litellm_llm_api_time_to_first_token_metric",
     "litellm_request_total_latency_metric",
     "litellm_overhead_latency_metric",
     "litellm_remaining_requests_metric",
@@ -173,9 +174,11 @@ DEFINED_PROMETHEUS_METRICS = Literal[
     "litellm_remaining_api_key_budget_metric",
     "litellm_api_key_max_budget_metric",
     "litellm_api_key_budget_remaining_hours_metric",
+    "litellm_deployment_state",
     "litellm_deployment_failure_responses",
     "litellm_deployment_total_requests",
     "litellm_deployment_success_responses",
+    "litellm_deployment_cooled_down",
     "litellm_pod_lock_manager_size",
     "litellm_in_memory_daily_spend_update_queue_size",
     "litellm_redis_daily_spend_update_queue_size",
@@ -191,9 +194,14 @@ class PrometheusMetricLabels:
         UserAPIKeyLabelNames.API_KEY_ALIAS.value,
         UserAPIKeyLabelNames.TEAM.value,
         UserAPIKeyLabelNames.TEAM_ALIAS.value,
-        UserAPIKeyLabelNames.REQUESTED_MODEL.value,
-        UserAPIKeyLabelNames.END_USER.value,
-        UserAPIKeyLabelNames.USER.value,
+    ]
+
+    litellm_llm_api_time_to_first_token_metric = [
+        UserAPIKeyLabelNames.v1_LITELLM_MODEL_NAME.value,
+        UserAPIKeyLabelNames.API_KEY_HASH.value,
+        UserAPIKeyLabelNames.API_KEY_ALIAS.value,
+        UserAPIKeyLabelNames.TEAM.value,
+        UserAPIKeyLabelNames.TEAM_ALIAS.value,
     ]
 
     litellm_request_total_latency_metric = [
@@ -315,6 +323,20 @@ class PrometheusMetricLabels:
         UserAPIKeyLabelNames.REQUESTED_MODEL.value,
     ]
 
+    litellm_deployment_state = [
+        UserAPIKeyLabelNames.v2_LITELLM_MODEL_NAME.value,
+        UserAPIKeyLabelNames.MODEL_ID.value,
+        UserAPIKeyLabelNames.API_BASE.value,
+        UserAPIKeyLabelNames.API_PROVIDER.value,
+    ]
+
+    litellm_deployment_cooled_down = [
+        UserAPIKeyLabelNames.v2_LITELLM_MODEL_NAME.value,
+        UserAPIKeyLabelNames.MODEL_ID.value,
+        UserAPIKeyLabelNames.API_BASE.value,
+        UserAPIKeyLabelNames.API_PROVIDER.value,
+    ]
+
     litellm_deployment_successful_fallbacks = [
         UserAPIKeyLabelNames.REQUESTED_MODEL.value,
         UserAPIKeyLabelNames.FALLBACK_MODEL.value,