fix: req changes + UI fixed

litellm/proxy/hooks/max_budget_per_session_limiter.py

@@ -129,7 +129,11 @@ class _PROXY_MaxBudgetPerSessionHandler(CustomLogger):
         """
         try:
             litellm_params = kwargs.get("litellm_params") or {}
-            metadata = litellm_params.get("metadata") or {}
+            metadata = (
+                litellm_params.get("metadata")
+                or litellm_params.get("litellm_metadata")
+                or {}
+            )
             session_id = metadata.get("session_id")
             if session_id is None:
                 return
@@ -208,9 +212,7 @@ class _PROXY_MaxBudgetPerSessionHandler(CustomLogger):
 
     async def _get_current_spend(self, cache_key: str) -> float:
         """Read current accumulated spend for a session."""
-        if (
-            self.internal_usage_cache.dual_cache.redis_cache is not None
-        ):
+        if self.internal_usage_cache.dual_cache.redis_cache is not None:
             try:
                 result = await self.internal_usage_cache.dual_cache.redis_cache.async_get_cache(
                     key=cache_key
@@ -252,9 +254,7 @@ class _PROXY_MaxBudgetPerSessionHandler(CustomLogger):
 
         return await self._in_memory_increment_spend(cache_key, amount)
 
-    async def _in_memory_increment_spend(
-        self, cache_key: str, amount: float
-    ) -> float:
+    async def _in_memory_increment_spend(self, cache_key: str, amount: float) -> float:
         current = await self.internal_usage_cache.async_get_cache(
             key=cache_key,
             litellm_parent_otel_span=None,

litellm/proxy/hooks/parallel_request_limiter_v3.py

@@ -932,6 +932,26 @@ class _PROXY_MaxParallelRequestsHandler_v3(CustomLogger):
                     )
                 )
 
+            # Session-level rate limits (per agent + session_id)
+            if _agent is not None and (
+                _agent.session_tpm_limit is not None
+                or _agent.session_rpm_limit is not None
+            ):
+                _metadata = data.get("metadata") or data.get("litellm_metadata") or {}
+                _session_id = _metadata.get("session_id")
+                if _session_id is not None:
+                    descriptors.append(
+                        RateLimitDescriptor(
+                            key="agent_session",
+                            value=f"{_agent_id}:{_session_id}",
+                            rate_limit={
+                                "requests_per_unit": _agent.session_rpm_limit,
+                                "tokens_per_unit": _agent.session_tpm_limit,
+                                "window_size": self.window_size,
+                            },
+                        )
+                    )
+
         # Model rate limits
         requested_model = data.get("model", None)
         self._add_model_per_key_rate_limit_descriptor(
@@ -1405,7 +1425,9 @@ class _PROXY_MaxParallelRequestsHandler_v3(CustomLogger):
             return "total"  # default to total
         return specified_rate_limit_type
 
-    async def async_log_success_event(self, kwargs, response_obj, start_time, end_time):
+    async def async_log_success_event(
+        self, kwargs, response_obj, start_time, end_time
+    ):  # noqa: PLR0915
         """
         Update TPM usage on successful API calls by incrementing counters using pipeline
         """
@@ -1526,7 +1548,7 @@ class _PROXY_MaxParallelRequestsHandler_v3(CustomLogger):
                     )
                 )
 
-            # Agent TPM
+            # Agent TPM + Agent session TPM
             user_api_key_agent_id = standard_logging_metadata.get(
                 "user_api_key_agent_id"
             )
@@ -1539,6 +1561,20 @@ class _PROXY_MaxParallelRequestsHandler_v3(CustomLogger):
                         total_tokens=total_tokens,
                     )
                 )
+                # Per-session token tracking (keyed by agent_id:session_id)
+                _lp = kwargs.get("litellm_params") or {}
+                _session_id = (
+                    _lp.get("metadata") or _lp.get("litellm_metadata") or {}
+                ).get("session_id")
+                if _session_id:
+                    pipeline_operations.extend(
+                        self._create_pipeline_operations(
+                            key="agent_session",
+                            value=f"{user_api_key_agent_id}:{_session_id}",
+                            rate_limit_type="tokens",
+                            total_tokens=total_tokens,
+                        )
+                    )
 
             # Model-specific TPM
             if model_group and user_api_key:

ui/litellm-dashboard/src/components/agents/agent_info.tsx

@@ -205,6 +205,15 @@ const AgentInfoView: React.FC<AgentInfoViewProps> = ({
               <Descriptions.Item label="RPM Limit">{agent.rpm_limit ?? "Unlimited"}</Descriptions.Item>
               <Descriptions.Item label="Session TPM Limit">{agent.session_tpm_limit ?? "Unlimited"}</Descriptions.Item>
               <Descriptions.Item label="Session RPM Limit">{agent.session_rpm_limit ?? "Unlimited"}</Descriptions.Item>
+              {agent.spend !== undefined && agent.spend !== null && (
+                <Descriptions.Item label="Total Spend">${agent.spend.toFixed(6)}</Descriptions.Item>
+              )}
+              {agent.litellm_params?.max_iterations !== undefined && (
+                <Descriptions.Item label="Max Iterations">{agent.litellm_params.max_iterations}</Descriptions.Item>
+              )}
+              {agent.litellm_params?.max_budget_per_session !== undefined && (
+                <Descriptions.Item label="Max Budget Per Session">${agent.litellm_params.max_budget_per_session}</Descriptions.Item>
+              )}
               <Descriptions.Item label="Created At">{formatDate(agent.created_at)}</Descriptions.Item>
               <Descriptions.Item label="Updated At">{formatDate(agent.updated_at)}</Descriptions.Item>
             </Descriptions>