scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-10-10 03:28:53 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 84

MCP - usage tracking (#12397)

Browse source 
* fix(common_daily_activity.py): initial commit with working mock BE endpoint for mcp usage

* feat(ui/): show mcp server activity on UI

allows admin to know which mcp's are being used

* feat(common_daily_activity.py): return activity by key

* feat(ui/): show top api keys for a given model / mcp server

allow user to know which key is driving spend

* fix(common_daily_activity.py): use known mcp server names

* feat(server.py): log the namespaced tool name (includes server prefix)

allow accurate cost tracking

* feat(db_spend_update_writer.py): log by mcp_namespaced_tool_name

store aggregate daily activity by mcp_namespaced_tool_name

Enables cost / usage tracking by mcp tool name

* fix(server.py): add key/user metadata to mcp calls

* refactor(common_daily_activity.py): update to return mcp activity in API

* fix(common_daily_activity.py): handle empty key

* fix(common_daily_activity.py): track when api key is empty

* test(test_spend_management_endpoints.py): update tests

* fix: fix ui linting error

* fix: fix linting errors

* test: add missing key

* build(schema.prisma): add mcp tool tracking

* fix(migration.sql): add schema migration file

* feat(server.py): add request logging for mcp calls

enables storing the mcp calls

* fix(new_usage.tsx): fix linting errors

* fix: fix code qa errors

* fix(activity_metrics.tsx): fix ui linting errors post-merge

* fix(types/utils.py): fix linting error

* fix(server.py): always have name
This commit is contained in:
Krish Dholakia 2025-07-08 22:08:16 -07:00 • committed by GitHub
parent e46a23db53
commit c4af2eb5e2
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
19 changed files with 689 additions and 208 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

42
litellm-proxy-extras/litellm_proxy_extras/migrations/20250707230009_add_mcp_namespaced_tool_name/migration.sql Normal file
View file

@ -0,0 +1,42 @@
-- DropIndex
DROP INDEX "LiteLLM_DailyTagSpend_tag_date_api_key_model_custom_llm_pro_key";
-- DropIndex
DROP INDEX "LiteLLM_DailyTeamSpend_team_id_date_api_key_model_custom_ll_key";
-- DropIndex
DROP INDEX "LiteLLM_DailyUserSpend_user_id_date_api_key_model_custom_ll_key";
-- AlterTable
ALTER TABLE "LiteLLM_DailyTagSpend" ADD COLUMN "mcp_namespaced_tool_name" TEXT,
ALTER COLUMN "model" DROP NOT NULL;
-- AlterTable
ALTER TABLE "LiteLLM_DailyTeamSpend" ADD COLUMN "mcp_namespaced_tool_name" TEXT,
ALTER COLUMN "model" DROP NOT NULL;
-- AlterTable
ALTER TABLE "LiteLLM_DailyUserSpend" ADD COLUMN "mcp_namespaced_tool_name" TEXT,
ALTER COLUMN "model" DROP NOT NULL;
-- AlterTable
ALTER TABLE "LiteLLM_SpendLogs" ADD COLUMN "mcp_namespaced_tool_name" TEXT;
-- CreateIndex
CREATE INDEX "LiteLLM_DailyTagSpend_mcp_namespaced_tool_name_idx" ON "LiteLLM_DailyTagSpend"("mcp_namespaced_tool_name");
-- CreateIndex
CREATE UNIQUE INDEX "LiteLLM_DailyTagSpend_tag_date_api_key_model_custom_llm_pro_key" ON "LiteLLM_DailyTagSpend"("tag", "date", "api_key", "model", "custom_llm_provider", "mcp_namespaced_tool_name");
-- CreateIndex
CREATE INDEX "LiteLLM_DailyTeamSpend_mcp_namespaced_tool_name_idx" ON "LiteLLM_DailyTeamSpend"("mcp_namespaced_tool_name");
-- CreateIndex
CREATE UNIQUE INDEX "LiteLLM_DailyTeamSpend_team_id_date_api_key_model_custom_ll_key" ON "LiteLLM_DailyTeamSpend"("team_id", "date", "api_key", "model", "custom_llm_provider", "mcp_namespaced_tool_name");
-- CreateIndex
CREATE INDEX "LiteLLM_DailyUserSpend_mcp_namespaced_tool_name_idx" ON "LiteLLM_DailyUserSpend"("mcp_namespaced_tool_name");
-- CreateIndex
CREATE UNIQUE INDEX "LiteLLM_DailyUserSpend_user_id_date_api_key_model_custom_ll_key" ON "LiteLLM_DailyUserSpend"("user_id", "date", "api_key", "model", "custom_llm_provider", "mcp_namespaced_tool_name");

25
litellm-proxy-extras/litellm_proxy_extras/schema.prisma
View file

@ -262,6 +262,7 @@ model LiteLLM_SpendLogs {
response Json? @default("{}")
session_id String?
status String?
mcp_namespaced_tool_name String?
proxy_server_request Json? @default("{}")
@@index([startTime])
@@index([end_user])
@ -360,9 +361,10 @@ model LiteLLM_DailyUserSpend {
user_id String?
date String
api_key String
model String
model String?
model_group String?
custom_llm_provider String?
custom_llm_provider String?
mcp_namespaced_tool_name String?
prompt_tokens BigInt @default(0)
completion_tokens BigInt @default(0)
cache_read_input_tokens BigInt @default(0)
@ -374,11 +376,12 @@ model LiteLLM_DailyUserSpend {
created_at DateTime @default(now())
updated_at DateTime @updatedAt
@@unique([user_id, date, api_key, model, custom_llm_provider])
@@unique([user_id, date, api_key, model, custom_llm_provider, mcp_namespaced_tool_name])
@@index([date])
@@index([user_id])
@@index([api_key])
@@index([model])
@@index([mcp_namespaced_tool_name])
}
// Track daily team spend metrics per model and key
@ -387,9 +390,10 @@ model LiteLLM_DailyTeamSpend {
team_id String?
date String
api_key String
model String
model String?
model_group String?
custom_llm_provider String?
custom_llm_provider String?
mcp_namespaced_tool_name String?
prompt_tokens BigInt @default(0)
completion_tokens BigInt @default(0)
cache_read_input_tokens BigInt @default(0)
@ -401,11 +405,12 @@ model LiteLLM_DailyTeamSpend {
created_at DateTime @default(now())
updated_at DateTime @updatedAt
@@unique([team_id, date, api_key, model, custom_llm_provider])
@@unique([team_id, date, api_key, model, custom_llm_provider, mcp_namespaced_tool_name])
@@index([date])
@@index([team_id])
@@index([api_key])
@@index([model])
@@index([mcp_namespaced_tool_name])
}
// Track daily team spend metrics per model and key
@ -414,9 +419,10 @@ model LiteLLM_DailyTagSpend {
tag String?
date String
api_key String
model String
model String?
model_group String?
custom_llm_provider String?
custom_llm_provider String?
mcp_namespaced_tool_name String?
prompt_tokens BigInt @default(0)
completion_tokens BigInt @default(0)
cache_read_input_tokens BigInt @default(0)
@ -428,11 +434,12 @@ model LiteLLM_DailyTagSpend {
created_at DateTime @default(now())
updated_at DateTime @updatedAt
@@unique([tag, date, api_key, model, custom_llm_provider])
@@unique([tag, date, api_key, model, custom_llm_provider, mcp_namespaced_tool_name])
@@index([date])
@@index([tag])
@@index([api_key])
@@index([model])
@@index([mcp_namespaced_tool_name])
}

122
litellm/proxy/_experimental/mcp_server/server.py
View file

@ -11,7 +11,6 @@ from pydantic import ConfigDict
from starlette.types import Receive, Scope, Send
from litellm._logging import verbose_logger
from litellm.constants import MCP_TOOL_NAME_PREFIX
from litellm.litellm_core_utils.litellm_logging import Logging as LiteLLMLoggingObj
from litellm.proxy._experimental.mcp_server.auth.user_api_key_auth_mcp import (
MCPRequestHandler,
@ -130,7 +129,9 @@ if MCP_AVAILABLE:
await _sse_session_manager_cm.__aenter__()
_SESSION_MANAGERS_INITIALIZED = True
verbose_logger.info("MCP Server started with StreamableHTTP and SSE session managers!")
verbose_logger.info(
"MCP Server started with StreamableHTTP and SSE session managers!"
)
async def shutdown_session_managers():
"""Shutdown the session managers."""
@ -198,17 +199,48 @@ if MCP_AVAILABLE:
Raises:
HTTPException: If tool not found or arguments missing
"""
from fastapi import Request
from litellm.proxy.litellm_pre_call_utils import add_litellm_data_to_request
from litellm.proxy.proxy_server import proxy_config
# Validate arguments
user_api_key_auth, mcp_auth_header, _ = get_auth_context()
verbose_logger.debug(
f"MCP mcp_server_tool_call - User API Key Auth from context: {user_api_key_auth}"
)
response = await call_mcp_tool(
name=name,
arguments=arguments,
user_api_key_auth=user_api_key_auth,
mcp_auth_header=mcp_auth_header,
)
try:
# Create a body date for logging
body_data = {"name": name, "arguments": arguments}
request = Request(
scope={
"type": "http",
"method": "POST",
"path": "/mcp/tools/call",
"headers": [(b"content-type", b"application/json")],
}
)
if user_api_key_auth is not None:
data = await add_litellm_data_to_request(
data=body_data,
request=request,
user_api_key_dict=user_api_key_auth,
proxy_config=proxy_config,
)
else:
data = body_data
response = await call_mcp_tool(
user_api_key_auth=user_api_key_auth,
mcp_auth_header=mcp_auth_header,
**data, # for logging
)
except Exception as e:
verbose_logger.exception(f"MCP mcp_server_tool_call - error: {e}")
raise e
return response
########################################################
@ -222,7 +254,7 @@ if MCP_AVAILABLE:
async def _get_tools_from_mcp_servers(
user_api_key_auth: Optional[UserAPIKeyAuth],
mcp_auth_header: Optional[str],
mcp_servers: Optional[List[str]]
mcp_servers: Optional[List[str]],
) -> List[MCPTool]:
"""
Helper method to fetch tools from MCP servers based on server filtering criteria.
@ -238,12 +270,19 @@ if MCP_AVAILABLE:
if mcp_servers:
# If mcp_servers header is present, only get tools from specified servers
tools = []
for server_id in await global_mcp_server_manager.get_allowed_mcp_servers(user_api_key_auth):
for server_id in await global_mcp_server_manager.get_allowed_mcp_servers(
user_api_key_auth
):
server = global_mcp_server_manager.get_mcp_server_by_id(server_id)
if server and any(normalize_server_name(server.name) == normalize_server_name(s) for s in mcp_servers):
server_tools = await global_mcp_server_manager._get_tools_from_server(
server=server,
mcp_auth_header=mcp_auth_header,
if server and any(
normalize_server_name(server.name) == normalize_server_name(s)
for s in mcp_servers
):
server_tools = (
await global_mcp_server_manager._get_tools_from_server(
server=server,
mcp_auth_header=mcp_auth_header,
)
)
tools.extend(server_tools)
return tools
@ -284,7 +323,7 @@ if MCP_AVAILABLE:
tools_from_mcp_servers = await _get_tools_from_mcp_servers(
user_api_key_auth=user_api_key_auth,
mcp_auth_header=mcp_auth_header,
mcp_servers=mcp_servers
mcp_servers=mcp_servers,
)
verbose_logger.debug("TOOLS FROM MCP SERVERS: %s", tools_from_mcp_servers)
@ -294,28 +333,31 @@ if MCP_AVAILABLE:
@client
async def call_mcp_tool(
name: str,
arguments: Optional[Dict[str, Any]] = None,
user_api_key_auth: Optional[UserAPIKeyAuth] = None,
mcp_auth_header: Optional[str] = None,
**kwargs: Any
name: str,
arguments: Optional[Dict[str, Any]] = None,
user_api_key_auth: Optional[UserAPIKeyAuth] = None,
mcp_auth_header: Optional[str] = None,
**kwargs: Any,
) -> List[Union[MCPTextContent, MCPImageContent, MCPEmbeddedResource]]:
"""
Call a specific tool with the provided arguments (handles prefixed tool names)
"""
if arguments is None:
raise HTTPException(
status_code=400, detail="Request arguments are required"
)
# Remove prefix from tool name for logging and processing
original_tool_name, _ = get_server_name_prefix_tool_mcp(
name)
original_tool_name, server_name_from_prefix = get_server_name_prefix_tool_mcp(
name
)
standard_logging_mcp_tool_call: StandardLoggingMCPToolCall = (
_get_standard_logging_mcp_tool_call(
name=original_tool_name, # Use original name for logging
arguments=arguments,
server_name=server_name_from_prefix,
)
)
litellm_logging_obj: Optional[LiteLLMLoggingObj] = kwargs.get(
@ -325,20 +367,16 @@ if MCP_AVAILABLE:
litellm_logging_obj.model_call_details["mcp_tool_call_metadata"] = (
standard_logging_mcp_tool_call
)
model_name = f"MCP: {MCP_TOOL_NAME_PREFIX}: {standard_logging_mcp_tool_call.get('name') or ''}"
litellm_logging_obj.model = model_name
litellm_logging_obj.model_call_details["model"] = model_name
litellm_logging_obj.model_call_details["custom_llm_provider"] = (
standard_logging_mcp_tool_call.get("mcp_server_name")
)
#########################################################
# Managed MCP Server Tool
# Try managed server tool first (pass the full prefixed name)
# Primary and recommended way to use MCP servers
#########################################################
mcp_server: Optional[MCPServer] = global_mcp_server_manager._get_mcp_server_from_tool_name(name)
mcp_server: Optional[MCPServer] = (
global_mcp_server_manager._get_mcp_server_from_tool_name(name)
)
if mcp_server:
standard_logging_mcp_tool_call["mcp_server_cost_info"] = (mcp_server.mcp_info or {}).get("mcp_server_cost_info")
standard_logging_mcp_tool_call["mcp_server_cost_info"] = (
mcp_server.mcp_info or {}
).get("mcp_server_cost_info")
return await _handle_managed_mcp_tool(
name=name, # Pass the full name (potentially prefixed)
arguments=arguments,
@ -355,6 +393,7 @@ if MCP_AVAILABLE:
def _get_standard_logging_mcp_tool_call(
name: str,
arguments: Dict[str, Any],
server_name: Optional[str],
) -> StandardLoggingMCPToolCall:
mcp_server = global_mcp_server_manager._get_mcp_server_from_tool_name(name)
if mcp_server:
@ -364,15 +403,17 @@ if MCP_AVAILABLE:
arguments=arguments,
mcp_server_name=mcp_info.get("server_name"),
mcp_server_logo_url=mcp_info.get("logo_url"),
namespaced_tool_name=f"{server_name}/{name}" if server_name else name,
)
else:
return StandardLoggingMCPToolCall(
name=name,
arguments=arguments,
namespaced_tool_name=f"{server_name}/{name}" if server_name else name,
)
async def _handle_managed_mcp_tool(
name: str,
name: str,
arguments: Dict[str, Any],
user_api_key_auth: Optional[UserAPIKeyAuth] = None,
mcp_auth_header: Optional[str] = None,
@ -388,7 +429,7 @@ if MCP_AVAILABLE:
return call_tool_result.content # type: ignore[return-value]
async def _handle_local_mcp_tool(
name: str, arguments: Dict[str, Any]
name: str, arguments: Dict[str, Any]
) -> List[Union[MCPTextContent, MCPImageContent, MCPEmbeddedResource]]:
"""
Handle tool execution for local registry tools
@ -404,7 +445,6 @@ if MCP_AVAILABLE:
except Exception as e:
return [MCPTextContent(text=f"Error: {str(e)}", type="text")]
async def handle_streamable_http_mcp(
scope: Scope, receive: Receive, send: Send
) -> None:
@ -486,7 +526,7 @@ if MCP_AVAILABLE:
########################################################
def set_auth_context(
user_api_key_auth: UserAPIKeyAuth,
user_api_key_auth: UserAPIKeyAuth,
mcp_auth_header: Optional[str] = None,
mcp_servers: Optional[List[str]] = None,
) -> None:
@ -505,7 +545,9 @@ if MCP_AVAILABLE:
)
auth_context_var.set(auth_user)
def get_auth_context() -> Tuple[Optional[UserAPIKeyAuth], Optional[str], Optional[List[str]]]:
def get_auth_context() -> (
Tuple[Optional[UserAPIKeyAuth], Optional[str], Optional[List[str]]]
):
"""
Get the UserAPIKeyAuth from the auth context variable.
@ -514,7 +556,11 @@ if MCP_AVAILABLE:
"""
auth_user = auth_context_var.get()
if auth_user and isinstance(auth_user, MCPAuthenticatedUser):
return auth_user.user_api_key_auth, auth_user.mcp_auth_header, auth_user.mcp_servers
return (
auth_user.user_api_key_auth,
auth_user.mcp_auth_header,
auth_user.mcp_servers,
)
return None, None, None
########################################################

4
litellm/proxy/_types.py
View file

@ -2176,6 +2176,7 @@ class SpendLogsPayload(TypedDict):
model: str
model_id: Optional[str]
model_group: Optional[str]
mcp_namespaced_tool_name: Optional[str]
api_base: str
user: str
metadata: str # json str
@ -3056,8 +3057,9 @@ class DefaultInternalUserParams(LiteLLMPydanticObjectBase):
class BaseDailySpendTransaction(TypedDict):
date: str
api_key: str
model: str
model: Optional[str]
model_group: Optional[str]
mcp_namespaced_tool_name: Optional[str]
custom_llm_provider: Optional[str]
# token count metrics

59
litellm/proxy/db/db_spend_update_writer.py
View file

@ -775,6 +775,8 @@ class DBSpendUpdateWriter:
e=e, start_time=start_time, proxy_logging_obj=proxy_logging_obj
)
# fmt: off
@overload
@staticmethod
async def _update_daily_spend(
@ -786,7 +788,7 @@ class DBSpendUpdateWriter:
entity_id_field: str,
table_name: str,
unique_constraint_name: str,
) -> None:
) -> None:
...
@overload
@ -814,8 +816,9 @@ class DBSpendUpdateWriter:
entity_id_field: str,
table_name: str,
unique_constraint_name: str,
) -> None:
) -> None:
...
# fmt: on
@staticmethod
async def _update_daily_spend(
@ -870,6 +873,9 @@ class DBSpendUpdateWriter:
"custom_llm_provider": transaction.get(
"custom_llm_provider"
),
"mcp_namespaced_tool_name": transaction.get(
"mcp_namespaced_tool_name"
),
}
}
@ -881,8 +887,11 @@ class DBSpendUpdateWriter:
entity_id_field: entity_id,
"date": transaction["date"],
"api_key": transaction["api_key"],
"model": transaction["model"],
"model": transaction.get("model"),
"model_group": transaction.get("model_group"),
"mcp_namespaced_tool_name": transaction.get(
"mcp_namespaced_tool_name"
),
"custom_llm_provider": transaction.get(
"custom_llm_provider"
),
@ -898,13 +907,13 @@ class DBSpendUpdateWriter:
# Add cache-related fields if they exist
if "cache_read_input_tokens" in transaction:
common_data[
"cache_read_input_tokens"
] = transaction.get("cache_read_input_tokens", 0)
common_data["cache_read_input_tokens"] = (
transaction.get("cache_read_input_tokens", 0)
)
if "cache_creation_input_tokens" in transaction:
common_data[
"cache_creation_input_tokens"
] = transaction.get("cache_creation_input_tokens", 0)
common_data["cache_creation_input_tokens"] = (
transaction.get("cache_creation_input_tokens", 0)
)
# Create update data structure
update_data = {
@ -993,7 +1002,7 @@ class DBSpendUpdateWriter:
entity_type="user",
entity_id_field="user_id",
table_name="litellm_dailyuserspend",
unique_constraint_name="user_id_date_api_key_model_custom_llm_provider",
unique_constraint_name="user_id_date_api_key_model_custom_llm_provider_mcp_namespaced_tool_name",
)
@staticmethod
@ -1014,7 +1023,7 @@ class DBSpendUpdateWriter:
entity_type="team",
entity_id_field="team_id",
table_name="litellm_dailyteamspend",
unique_constraint_name="team_id_date_api_key_model_custom_llm_provider",
unique_constraint_name="team_id_date_api_key_model_custom_llm_provider_mcp_namespaced_tool_name",
)
@staticmethod
@ -1035,7 +1044,7 @@ class DBSpendUpdateWriter:
entity_type="tag",
entity_id_field="tag",
table_name="litellm_dailytagspend",
unique_constraint_name="tag_date_api_key_model_custom_llm_provider",
unique_constraint_name="tag_date_api_key_model_custom_llm_provider_mcp_namespaced_tool_name",
)
async def _common_add_spend_log_transaction_to_daily_transaction(
@ -1044,7 +1053,7 @@ class DBSpendUpdateWriter:
prisma_client: PrismaClient,
type: Literal["user", "team", "request_tags"] = "user",
) -> Optional[BaseDailySpendTransaction]:
common_expected_keys = ["startTime", "api_key", "model", "custom_llm_provider"]
common_expected_keys = ["startTime", "api_key"]
if type == "user":
expected_keys = ["user", *common_expected_keys]
elif type == "team":
@ -1053,13 +1062,28 @@ class DBSpendUpdateWriter:
expected_keys = ["request_tags", *common_expected_keys]
else:
raise ValueError(f"Invalid type: {type}")
if not all(key in payload for key in expected_keys):
verbose_proxy_logger.debug(
f"Missing expected keys: {expected_keys}, in payload, skipping from daily_user_spend_transactions"
)
return None
any_expected_keys = ["model", "mcp_namespaced_tool_name"]
if not any(key in payload for key in any_expected_keys):
verbose_proxy_logger.debug(
f"Missing any expected keys: {any_expected_keys}, in payload, skipping from daily_user_spend_transactions"
)
return None
elif "mcp_namespaced_tool_name" in payload:
pass
elif "model" in payload and (
"custom_llm_provider" not in payload or "model_group" not in payload
):
verbose_proxy_logger.debug(
"Missing custom_llm_provider or model_group in payload, skipping from daily_user_spend_transactions"
)
return None
request_status = prisma_client.get_request_status(payload)
verbose_proxy_logger.info(f"Logged request status: {request_status}")
_metadata: SpendLogsMetadata = json.loads(payload["metadata"])
@ -1078,9 +1102,10 @@ class DBSpendUpdateWriter:
daily_transaction = BaseDailySpendTransaction(
date=date,
api_key=payload["api_key"],
model=payload["model"],
model_group=payload["model_group"],
custom_llm_provider=payload["custom_llm_provider"],
model=payload.get("model", None),
model_group=payload.get("model_group", None),
mcp_namespaced_tool_name=payload.get("mcp_namespaced_tool_name", None),
custom_llm_provider=payload.get("custom_llm_provider", None),
prompt_tokens=payload["prompt_tokens"],
completion_tokens=payload["completion_tokens"],
spend=payload["spend"],

144
litellm/proxy/management_endpoints/common_daily_activity.py
View file

@ -44,16 +44,81 @@ def update_breakdown_metrics(
"""Updates breakdown metrics for a single record using the existing update_metrics function"""
# Update model breakdown
if record.model not in breakdown.models:
if record.model and record.model not in breakdown.models:
breakdown.models[record.model] = MetricWithMetadata(
metrics=SpendMetrics(),
metadata=model_metadata.get(
record.model, {}
), # Add any model-specific metadata here
)
breakdown.models[record.model].metrics = update_metrics(
breakdown.models[record.model].metrics, record
)
if record.model:
breakdown.models[record.model].metrics = update_metrics(
breakdown.models[record.model].metrics, record
)
# Update API key breakdown for this model
if record.api_key not in breakdown.models[record.model].api_key_breakdown:
breakdown.models[record.model].api_key_breakdown[record.api_key] = (
KeyMetricWithMetadata(
metrics=SpendMetrics(),
metadata=KeyMetadata(
key_alias=api_key_metadata.get(record.api_key, {}).get(
"key_alias", None
),
team_id=api_key_metadata.get(record.api_key, {}).get(
"team_id", None
),
),
)
)
breakdown.models[record.model].api_key_breakdown[record.api_key].metrics = (
update_metrics(
breakdown.models[record.model]
.api_key_breakdown[record.api_key]
.metrics,
record,
)
)
if record.mcp_namespaced_tool_name:
if record.mcp_namespaced_tool_name not in breakdown.mcp_servers:
breakdown.mcp_servers[record.mcp_namespaced_tool_name] = MetricWithMetadata(
metrics=SpendMetrics(),
metadata={},
)
breakdown.mcp_servers[record.mcp_namespaced_tool_name].metrics = update_metrics(
breakdown.mcp_servers[record.mcp_namespaced_tool_name].metrics, record
)
# Update API key breakdown for this MCP server
if (
record.api_key
not in breakdown.mcp_servers[
record.mcp_namespaced_tool_name
].api_key_breakdown
):
breakdown.mcp_servers[record.mcp_namespaced_tool_name].api_key_breakdown[
record.api_key
] = KeyMetricWithMetadata(
metrics=SpendMetrics(),
metadata=KeyMetadata(
key_alias=api_key_metadata.get(record.api_key, {}).get(
"key_alias", None
),
team_id=api_key_metadata.get(record.api_key, {}).get(
"team_id", None
),
),
)
breakdown.mcp_servers[record.mcp_namespaced_tool_name].api_key_breakdown[
record.api_key
].metrics = update_metrics(
breakdown.mcp_servers[record.mcp_namespaced_tool_name]
.api_key_breakdown[record.api_key]
.metrics,
record,
)
# Update provider breakdown
provider = record.custom_llm_provider or "unknown"
@ -68,6 +133,28 @@ def update_breakdown_metrics(
breakdown.providers[provider].metrics, record
)
# Update API key breakdown for this provider
if record.api_key not in breakdown.providers[provider].api_key_breakdown:
breakdown.providers[provider].api_key_breakdown[record.api_key] = (
KeyMetricWithMetadata(
metrics=SpendMetrics(),
metadata=KeyMetadata(
key_alias=api_key_metadata.get(record.api_key, {}).get(
"key_alias", None
),
team_id=api_key_metadata.get(record.api_key, {}).get(
"team_id", None
),
),
)
)
breakdown.providers[provider].api_key_breakdown[record.api_key].metrics = (
update_metrics(
breakdown.providers[provider].api_key_breakdown[record.api_key].metrics,
record,
)
)
# Update api key breakdown
if record.api_key not in breakdown.api_keys:
breakdown.api_keys[record.api_key] = KeyMetricWithMetadata(
@ -92,14 +179,40 @@ def update_breakdown_metrics(
if entity_value not in breakdown.entities:
breakdown.entities[entity_value] = MetricWithMetadata(
metrics=SpendMetrics(),
metadata=entity_metadata_field.get(entity_value, {})
if entity_metadata_field
else {},
metadata=(
entity_metadata_field.get(entity_value, {})
if entity_metadata_field
else {}
),
)
breakdown.entities[entity_value].metrics = update_metrics(
breakdown.entities[entity_value].metrics, record
)
# Update API key breakdown for this entity
if record.api_key not in breakdown.entities[entity_value].api_key_breakdown:
breakdown.entities[entity_value].api_key_breakdown[record.api_key] = (
KeyMetricWithMetadata(
metrics=SpendMetrics(),
metadata=KeyMetadata(
key_alias=api_key_metadata.get(record.api_key, {}).get(
"key_alias", None
),
team_id=api_key_metadata.get(record.api_key, {}).get(
"team_id", None
),
),
)
)
breakdown.entities[entity_value].api_key_breakdown[record.api_key].metrics = (
update_metrics(
breakdown.entities[entity_value]
.api_key_breakdown[record.api_key]
.metrics,
record,
)
)
return breakdown
@ -131,6 +244,7 @@ async def get_daily_activity(
exclude_entity_ids: Optional[List[str]] = None,
) -> SpendAnalyticsPaginatedResponse:
"""Common function to get daily activity for any entity type."""
if prisma_client is None:
raise HTTPException(
status_code=500,
@ -181,6 +295,22 @@ async def get_daily_activity(
take=page_size,
)
# # for 50% of the records, set the mcp_server_id to a random value
# mcp_server_dict = {"Zapier_Gmail_MCP", "Stripe_MCP"}
# import random
# for idx, record in enumerate(daily_spend_data):
# record = LiteLLM_DailyUserSpend(**record.model_dump())
# if random.random() < 0.5:
# record.mcp_server_id = random.choice(list(mcp_server_dict))
# record.model = None
# record.model_group = None
# record.prompt_tokens = 0
# record.completion_tokens = 0
# record.cache_read_input_tokens = 0
# record.cache_creation_input_tokens = 0
# daily_spend_data[idx] = record
# Get all unique API keys from the spend data
api_keys = set()
for record in daily_spend_data:

19
litellm/proxy/management_endpoints/internal_user_endpoints.py
View file

@ -1451,16 +1451,17 @@ def update_breakdown_metrics(
"""Updates breakdown metrics for a single record using the existing update_metrics function"""
# Update model breakdown
if record.model not in breakdown.models:
breakdown.models[record.model] = MetricWithMetadata(
metrics=SpendMetrics(),
metadata=model_metadata.get(
record.model, {}
), # Add any model-specific metadata here
if record.model:
if record.model not in breakdown.models:
breakdown.models[record.model] = MetricWithMetadata(
metrics=SpendMetrics(),
metadata=model_metadata.get(
record.model, {}
), # Add any model-specific metadata here
)
breakdown.models[record.model].metrics = update_metrics(
breakdown.models[record.model].metrics, record
)
breakdown.models[record.model].metrics = update_metrics(
breakdown.models[record.model].metrics, record
)
# Update provider breakdown
provider = record.custom_llm_provider or "unknown"

25
litellm/proxy/schema.prisma
View file

@ -262,6 +262,7 @@ model LiteLLM_SpendLogs {
response Json? @default("{}")
session_id String?
status String?
mcp_namespaced_tool_name String?
proxy_server_request Json? @default("{}")
@@index([startTime])
@@index([end_user])
@ -360,9 +361,10 @@ model LiteLLM_DailyUserSpend {
user_id String?
date String
api_key String
model String
model String?
model_group String?
custom_llm_provider String?
custom_llm_provider String?
mcp_namespaced_tool_name String?
prompt_tokens BigInt @default(0)
completion_tokens BigInt @default(0)
cache_read_input_tokens BigInt @default(0)
@ -374,11 +376,12 @@ model LiteLLM_DailyUserSpend {
created_at DateTime @default(now())
updated_at DateTime @updatedAt
@@unique([user_id, date, api_key, model, custom_llm_provider])
@@unique([user_id, date, api_key, model, custom_llm_provider, mcp_namespaced_tool_name])
@@index([date])
@@index([user_id])
@@index([api_key])
@@index([model])
@@index([mcp_namespaced_tool_name])
}
// Track daily team spend metrics per model and key
@ -387,9 +390,10 @@ model LiteLLM_DailyTeamSpend {
team_id String?
date String
api_key String
model String
model String?
model_group String?
custom_llm_provider String?
custom_llm_provider String?
mcp_namespaced_tool_name String?
prompt_tokens BigInt @default(0)
completion_tokens BigInt @default(0)
cache_read_input_tokens BigInt @default(0)
@ -401,11 +405,12 @@ model LiteLLM_DailyTeamSpend {
created_at DateTime @default(now())
updated_at DateTime @updatedAt
@@unique([team_id, date, api_key, model, custom_llm_provider])
@@unique([team_id, date, api_key, model, custom_llm_provider, mcp_namespaced_tool_name])
@@index([date])
@@index([team_id])
@@index([api_key])
@@index([model])
@@index([mcp_namespaced_tool_name])
}
// Track daily team spend metrics per model and key
@ -414,9 +419,10 @@ model LiteLLM_DailyTagSpend {
tag String?
date String
api_key String
model String
model String?
model_group String?
custom_llm_provider String?
custom_llm_provider String?
mcp_namespaced_tool_name String?
prompt_tokens BigInt @default(0)
completion_tokens BigInt @default(0)
cache_read_input_tokens BigInt @default(0)
@ -428,11 +434,12 @@ model LiteLLM_DailyTagSpend {
created_at DateTime @default(now())
updated_at DateTime @updatedAt
@@unique([tag, date, api_key, model, custom_llm_provider])
@@unique([tag, date, api_key, model, custom_llm_provider, mcp_namespaced_tool_name])
@@index([date])
@@index([tag])
@@index([api_key])
@@index([model])
@@index([mcp_namespaced_tool_name])
}

8
litellm/proxy/spend_tracking/spend_tracking_utils.py
View file

@ -286,6 +286,13 @@ def get_logging_payload( # noqa: PLR0915
id = f"{id}_cache_hit{time.time()}" # SpendLogs does not allow duplicate request_id
mcp_namespaced_tool_name = None
mcp_tool_call_metadata = clean_metadata.get("mcp_tool_call_metadata", {})
if mcp_tool_call_metadata is not None:
mcp_namespaced_tool_name = mcp_tool_call_metadata.get(
"namespaced_tool_name", None
)
try:
payload: SpendLogsPayload = SpendLogsPayload(
request_id=str(id),
@ -311,6 +318,7 @@ def get_logging_payload( # noqa: PLR0915
api_base=litellm_params.get("api_base", ""),
model_group=_model_group,
model_id=_model_id,
mcp_namespaced_tool_name=mcp_namespaced_tool_name,
requester_ip_address=clean_metadata.get("requester_ip_address", None),
custom_llm_provider=kwargs.get("custom_llm_provider", ""),
messages=_get_messages_for_spend_logs_payload(

18
litellm/types/proxy/management_endpoints/common_daily_activity.py
View file

@ -31,10 +31,6 @@ class MetricBase(BaseModel):
metrics: SpendMetrics
class MetricWithMetadata(MetricBase):
metadata: Dict[str, Any] = Field(default_factory=dict)
class KeyMetadata(BaseModel):
"""Metadata for a key"""
@ -48,9 +44,20 @@ class KeyMetricWithMetadata(MetricBase):
metadata: KeyMetadata = Field(default_factory=KeyMetadata)
class MetricWithMetadata(MetricBase):
metadata: Dict[str, Any] = Field(default_factory=dict)
# API key breakdown for this metric (e.g., which API keys are using this MCP server)
api_key_breakdown: Dict[str, KeyMetricWithMetadata] = Field(
default_factory=dict
) # api_key -> {metrics, metadata}
class BreakdownMetrics(BaseModel):
"""Breakdown of spend by different dimensions"""
mcp_servers: Dict[str, MetricWithMetadata] = Field(
default_factory=dict
) # mcp_server -> {metrics, metadata}
models: Dict[str, MetricWithMetadata] = Field(
default_factory=dict
) # model -> {metrics, metadata}
@ -96,7 +103,8 @@ class LiteLLM_DailyUserSpend(BaseModel):
user_id: str
date: str
api_key: str
model: str
mcp_server_id: Optional[str] = None
model: Optional[str] = None
model_group: Optional[str] = None
custom_llm_provider: Optional[str] = None
prompt_tokens: int = 0

9
litellm/types/utils.py
View file

@ -1821,7 +1821,7 @@ class StandardLoggingUserAPIKeyMetadata(TypedDict):
user_api_key_request_route: Optional[str]
class StandardLoggingMCPToolCall(TypedDict, total=False):
class StandardLoggingMCPToolCall(TypedDict, total=False):
name: str
"""
Name of the tool to call
@ -1847,6 +1847,13 @@ class StandardLoggingMCPToolCall(TypedDict, total=False):
(this is to render the logo on the logs page on litellm ui)
"""
namespaced_tool_name: Optional[str]
"""
Namespaced tool name of the MCP tool that the tool call was made to
Includes the server name prefix if it exists - eg. `deepwiki-mcp/get_page_content`
"""
mcp_server_cost_info: Optional[MCPServerCostInfo]
"""
Cost per query for the MCP server tool call

25
schema.prisma
View file

@ -262,6 +262,7 @@ model LiteLLM_SpendLogs {
response Json? @default("{}")
session_id String?
status String?
mcp_namespaced_tool_name String?
proxy_server_request Json? @default("{}")
@@index([startTime])
@@index([end_user])
@ -360,9 +361,10 @@ model LiteLLM_DailyUserSpend {
user_id String?
date String
api_key String
model String
model String?
model_group String?
custom_llm_provider String?
custom_llm_provider String?
mcp_namespaced_tool_name String?
prompt_tokens BigInt @default(0)
completion_tokens BigInt @default(0)
cache_read_input_tokens BigInt @default(0)
@ -374,11 +376,12 @@ model LiteLLM_DailyUserSpend {
created_at DateTime @default(now())
updated_at DateTime @updatedAt
@@unique([user_id, date, api_key, model, custom_llm_provider])
@@unique([user_id, date, api_key, model, custom_llm_provider, mcp_namespaced_tool_name])
@@index([date])
@@index([user_id])
@@index([api_key])
@@index([model])
@@index([mcp_namespaced_tool_name])
}
// Track daily team spend metrics per model and key
@ -387,9 +390,10 @@ model LiteLLM_DailyTeamSpend {
team_id String?
date String
api_key String
model String
model String?
model_group String?
custom_llm_provider String?
custom_llm_provider String?
mcp_namespaced_tool_name String?
prompt_tokens BigInt @default(0)
completion_tokens BigInt @default(0)
cache_read_input_tokens BigInt @default(0)
@ -401,11 +405,12 @@ model LiteLLM_DailyTeamSpend {
created_at DateTime @default(now())
updated_at DateTime @updatedAt
@@unique([team_id, date, api_key, model, custom_llm_provider])
@@unique([team_id, date, api_key, model, custom_llm_provider, mcp_namespaced_tool_name])
@@index([date])
@@index([team_id])
@@index([api_key])
@@index([model])
@@index([mcp_namespaced_tool_name])
}
// Track daily team spend metrics per model and key
@ -414,9 +419,10 @@ model LiteLLM_DailyTagSpend {
tag String?
date String
api_key String
model String
model String?
model_group String?
custom_llm_provider String?
custom_llm_provider String?
mcp_namespaced_tool_name String?
prompt_tokens BigInt @default(0)
completion_tokens BigInt @default(0)
cache_read_input_tokens BigInt @default(0)
@ -428,11 +434,12 @@ model LiteLLM_DailyTagSpend {
created_at DateTime @default(now())
updated_at DateTime @updatedAt
@@unique([tag, date, api_key, model, custom_llm_provider])
@@unique([tag, date, api_key, model, custom_llm_provider, mcp_namespaced_tool_name])
@@index([date])
@@index([tag])
@@index([api_key])
@@index([model])
@@index([mcp_namespaced_tool_name])
}

3
tests/logging_callback_tests/gcs_pub_sub_body/spend_logs_payload.json
View file

@ -26,5 +26,6 @@
"messages": "{}",
"response": "{}",
"proxy_server_request": "{}",
"status": "success"
"status": "success",
"mcp_namespaced_tool_name": null
}

131
tests/test_litellm/proxy/_experimental/mcp_server/test_mcp_server.py Normal file
View file

@ -0,0 +1,131 @@
import asyncio
from unittest.mock import AsyncMock, MagicMock, patch
import pytest
from litellm.proxy._types import UserAPIKeyAuth
@pytest.mark.asyncio
async def test_mcp_server_tool_call_body_contains_request_data():
"""Test that proxy_server_request body contains name and arguments"""
try:
from litellm.proxy._experimental.mcp_server.server import (
mcp_server_tool_call,
set_auth_context,
)
except ImportError:
pytest.skip("MCP server not available")
# Setup test data
tool_name = "test_tool"
tool_arguments = {"param1": "value1", "param2": 123}
# Mock user auth
user_api_key_auth = UserAPIKeyAuth(api_key="test_key", user_id="test_user")
set_auth_context(user_api_key_auth)
# Mock the add_litellm_data_to_request function to capture the data
captured_data = {}
async def mock_add_litellm_data_to_request(
data, request, user_api_key_dict, proxy_config
):
captured_data.update(data)
# Simulate the proxy_server_request creation
captured_data["proxy_server_request"] = {
"url": str(request.url),
"method": request.method,
"headers": {},
"body": data.copy(), # This is what we want to test
}
return captured_data
# Mock the call_mcp_tool function to avoid actual tool execution
async def mock_call_mcp_tool(*args, **kwargs):
return [{"type": "text", "text": "mocked response"}]
with patch(
"litellm.proxy.litellm_pre_call_utils.add_litellm_data_to_request",
mock_add_litellm_data_to_request,
):
with patch(
"litellm.proxy._experimental.mcp_server.server.call_mcp_tool",
mock_call_mcp_tool,
):
with patch(
"litellm.proxy.proxy_server.proxy_config",
MagicMock(),
):
# Call the function
await mcp_server_tool_call(tool_name, tool_arguments)
# Verify the body contains the expected data
assert "proxy_server_request" in captured_data
assert "body" in captured_data["proxy_server_request"]
body = captured_data["proxy_server_request"]["body"]
assert body["name"] == tool_name
assert body["arguments"] == tool_arguments
@pytest.mark.asyncio
async def test_mcp_server_tool_call_body_with_none_arguments():
"""Test that proxy_server_request body handles None arguments correctly"""
try:
from litellm.proxy._experimental.mcp_server.server import (
mcp_server_tool_call,
set_auth_context,
)
except ImportError:
pytest.skip("MCP server not available")
# Setup test data
tool_name = "test_tool_no_args"
tool_arguments = None
# Mock user auth
user_api_key_auth = UserAPIKeyAuth(api_key="test_key", user_id="test_user")
set_auth_context(user_api_key_auth)
# Mock the add_litellm_data_to_request function to capture the data
captured_data = {}
async def mock_add_litellm_data_to_request(
data, request, user_api_key_dict, proxy_config
):
captured_data.update(data)
captured_data["proxy_server_request"] = {
"url": str(request.url),
"method": request.method,
"headers": {},
"body": data.copy(),
}
return captured_data
# Mock the call_mcp_tool function
async def mock_call_mcp_tool(*args, **kwargs):
return [{"type": "text", "text": "mocked response"}]
with patch(
"litellm.proxy.litellm_pre_call_utils.add_litellm_data_to_request",
mock_add_litellm_data_to_request,
):
with patch(
"litellm.proxy._experimental.mcp_server.server.call_mcp_tool",
mock_call_mcp_tool,
):
with patch(
"litellm.proxy.proxy_server.proxy_config",
MagicMock(),
):
# Call the function
await mcp_server_tool_call(tool_name, tool_arguments)
# Verify the body contains the expected data
assert "proxy_server_request" in captured_data
assert "body" in captured_data["proxy_server_request"]
body = captured_data["proxy_server_request"]["body"]
assert body["name"] == tool_name
assert body["arguments"] == tool_arguments # Should be None

54
tests/test_litellm/proxy/spend_tracking/test_spend_management_endpoints.py
View file

@ -763,6 +763,7 @@ class TestSpendLogsPayload:
"response": "{}",
"proxy_server_request": "{}",
"status": "success",
"mcp_namespaced_tool_name": None,
}
)
@ -855,6 +856,7 @@ class TestSpendLogsPayload:
"response": "{}",
"proxy_server_request": "{}",
"status": "success",
"mcp_namespaced_tool_name": None,
}
)
@ -945,9 +947,13 @@ class TestSpendLogsPayload:
"response": "{}",
"proxy_server_request": "{}",
"status": "success",
"mcp_namespaced_tool_name": None,
}
)
print(f"payload: {payload}")
print(f"expected_payload: {expected_payload}")
differences = _compare_nested_dicts(
payload, expected_payload, ignore_keys=ignored_keys
)
@ -1085,8 +1091,8 @@ async def test_global_spend_keys_endpoint_limit_validation(client, monkeypatch):
async def test_view_spend_logs_summarize_parameter(client, monkeypatch):
"""Test the new summarize parameter in the /spend/logs endpoint"""
import datetime
from datetime import timezone, timedelta
from datetime import timedelta, timezone
# Mock spend logs data
mock_spend_logs = [
{
@ -1096,7 +1102,9 @@ async def test_view_spend_logs_summarize_parameter(client, monkeypatch):
"user": "test_user_1",
"team_id": "team1",
"spend": 0.05,
"startTime": (datetime.datetime.now(timezone.utc) - timedelta(days=1)).isoformat(),
"startTime": (
datetime.datetime.now(timezone.utc) - timedelta(days=1)
).isoformat(),
"model": "gpt-3.5-turbo",
"prompt_tokens": 100,
"completion_tokens": 50,
@ -1109,23 +1117,25 @@ async def test_view_spend_logs_summarize_parameter(client, monkeypatch):
"user": "test_user_1",
"team_id": "team1",
"spend": 0.10,
"startTime": (datetime.datetime.now(timezone.utc) - timedelta(days=1)).isoformat(),
"startTime": (
datetime.datetime.now(timezone.utc) - timedelta(days=1)
).isoformat(),
"model": "gpt-4",
"prompt_tokens": 200,
"completion_tokens": 100,
"total_tokens": 300,
},
]
# Mock for unsummarized data (summarize=false)
class MockDB:
def __init__(self):
self.litellm_spendlogs = self
async def find_many(self, *args, **kwargs):
# Return individual log entries when summarize=false
return mock_spend_logs
async def group_by(self, *args, **kwargs):
# Return grouped data when summarize=true
# Simplified mock response for grouped data
@ -1134,17 +1144,17 @@ async def test_view_spend_logs_summarize_parameter(client, monkeypatch):
{
"api_key": "sk-test-key",
"user": "test_user_1",
"model": "gpt-3.5-turbo",
"model": "gpt-3.5-turbo",
"startTime": yesterday.strftime("%Y-%m-%dT%H:%M:%S.%fZ"),
"_sum": {"spend": 0.05}
"_sum": {"spend": 0.05},
},
{
"api_key": "sk-test-key",
"user": "test_user_1",
"user": "test_user_1",
"model": "gpt-4",
"startTime": yesterday.strftime("%Y-%m-%dT%H:%M:%S.%fZ"),
"_sum": {"spend": 0.10}
}
"_sum": {"spend": 0.10},
},
]
class MockPrismaClient:
@ -1156,7 +1166,9 @@ async def test_view_spend_logs_summarize_parameter(client, monkeypatch):
monkeypatch.setattr("litellm.proxy.proxy_server.prisma_client", mock_prisma_client)
# Set up test dates
start_date = (datetime.datetime.now(timezone.utc) - timedelta(days=2)).strftime("%Y-%m-%d")
start_date = (datetime.datetime.now(timezone.utc) - timedelta(days=2)).strftime(
"%Y-%m-%d"
)
end_date = datetime.datetime.now(timezone.utc).strftime("%Y-%m-%d")
# Test 1: summarize=false should return individual log entries
@ -1169,10 +1181,10 @@ async def test_view_spend_logs_summarize_parameter(client, monkeypatch):
},
headers={"Authorization": "Bearer sk-test"},
)
assert response.status_code == 200
data = response.json()
# Should return the raw log entries
assert isinstance(data, list)
assert len(data) == 2
@ -1180,7 +1192,7 @@ async def test_view_spend_logs_summarize_parameter(client, monkeypatch):
assert data[1]["id"] == "log2"
assert data[0]["request_id"] == "req1"
assert data[1]["request_id"] == "req2"
# Test 2: summarize=true should return grouped data
response = client.get(
"/spend/logs",
@ -1191,10 +1203,10 @@ async def test_view_spend_logs_summarize_parameter(client, monkeypatch):
},
headers={"Authorization": "Bearer sk-test"},
)
assert response.status_code == 200
data = response.json()
# Should return grouped/summarized data
assert isinstance(data, list)
# The structure should be different - grouped by date with aggregated spend
@ -1202,7 +1214,7 @@ async def test_view_spend_logs_summarize_parameter(client, monkeypatch):
assert "spend" in data[0]
assert "users" in data[0]
assert "models" in data[0]
# Test 3: default behavior (no summarize parameter) should maintain backward compatibility
response = client.get(
"/spend/logs",
@ -1212,10 +1224,10 @@ async def test_view_spend_logs_summarize_parameter(client, monkeypatch):
},
headers={"Authorization": "Bearer sk-test"},
)
assert response.status_code == 200
data = response.json()
# Should return grouped/summarized data (same as summarize=true)
assert isinstance(data, list)
assert "startTime" in data[0]

94
ui/litellm-dashboard/src/components/activity_metrics.tsx
View file

@ -1,20 +1,10 @@
import React from "react";
import {
Card,
Grid,
Text,
Title,
} from "@tremor/react";
import { AreaChart, BarChart } from "@tremor/react";
import React from 'react';
import { Card, Grid, Text, Title, Accordion, AccordionHeader, AccordionBody } from '@tremor/react';
import { AreaChart, BarChart } from '@tremor/react';
import { SpendMetrics, DailyData, ModelActivityData, MetricWithMetadata, KeyMetricWithMetadata, TopApiKeyData } from './usage/types';
import { Collapse } from 'antd';
import { formatNumberWithCommas } from '@/utils/dataUtils';
import type { CustomTooltipProps } from "@tremor/react";
import {
SpendMetrics,
DailyData,
ModelActivityData,
KeyMetricWithMetadata,
} from "./usage/types";
import { Collapse } from "antd";
import { formatNumberWithCommas } from "@/utils/dataUtils";
import {
valueFormatter,
valueFormatterSpend,
@ -191,6 +181,35 @@ const ModelSection = ({
</Card>
</Grid>
{/* Top API Keys Section */}
{metrics.top_api_keys && metrics.top_api_keys.length > 0 && (
<Card className="mt-4">
<Title>Top API Keys by Spend</Title>
<div className="mt-3">
<div className="grid grid-cols-1 gap-2">
{metrics.top_api_keys.map((keyData, index) => (
<div key={keyData.api_key} className="flex justify-between items-center p-3 bg-gray-50 rounded-lg">
<div>
<Text className="font-medium">
{keyData.key_alias || `${keyData.api_key.substring(0, 10)}...`}
</Text>
{keyData.team_id && (
<Text className="text-xs text-gray-500">Team: {keyData.team_id}</Text>
)}
</div>
<div className="text-right">
<Text className="font-medium">${formatNumberWithCommas(keyData.spend, 2)}</Text>
<Text className="text-xs text-gray-500">
{keyData.requests.toLocaleString()} requests | {keyData.tokens.toLocaleString()} tokens
</Text>
</div>
</div>
))}
</div>
</div>
</Card>
)}
{/* Charts */}
<Grid numItems={2} className="gap-4">
<Card>
@ -542,10 +561,7 @@ const formatKeyLabel = (
};
// Process data function
export const processActivityData = (
dailyActivity: { results: DailyData[] },
key: "models" | "api_keys"
): Record<string, ModelActivityData> => {
export const processActivityData = (dailyActivity: { results: DailyData[] }, key: "models" | "api_keys" | "mcp_servers"): Record<string, ModelActivityData> => {
const modelMetrics: Record<string, ModelActivityData> = {};
dailyActivity.results.forEach((day) => {
@ -565,7 +581,8 @@ export const processActivityData = (
total_spend: 0,
total_cache_read_input_tokens: 0,
total_cache_creation_input_tokens: 0,
daily_data: [],
top_api_keys: [],
daily_data: []
};
}
@ -605,6 +622,41 @@ export const processActivityData = (
});
});
// Process API key breakdowns for each metric (skip if key is 'api_keys' to avoid duplication)
if (key !== 'api_keys') {
Object.entries(modelMetrics).forEach(([model, _]) => {
const apiKeyBreakdown: Record<string, TopApiKeyData> = {};
// Aggregate API key data across all days
dailyActivity.results.forEach((day) => {
const modelData = day.breakdown[key]?.[model];
if (modelData && 'api_key_breakdown' in modelData) {
Object.entries(modelData.api_key_breakdown || {}).forEach(([apiKey, keyData]) => {
if (!apiKeyBreakdown[apiKey]) {
apiKeyBreakdown[apiKey] = {
api_key: apiKey,
key_alias: keyData.metadata.key_alias,
team_id: keyData.metadata.team_id,
spend: 0,
requests: 0,
tokens: 0,
};
}
apiKeyBreakdown[apiKey].spend += keyData.metrics.spend;
apiKeyBreakdown[apiKey].requests += keyData.metrics.api_requests;
apiKeyBreakdown[apiKey].tokens += keyData.metrics.total_tokens;
});
}
});
// Sort by spend and take top 5
modelMetrics[model].top_api_keys = Object.values(apiKeyBreakdown)
.sort((a, b) => b.spend - a.spend)
.slice(0, 5);
});
}
// Sort daily data
Object.values(modelMetrics).forEach((metrics) => {
metrics.daily_data.sort(

28
ui/litellm-dashboard/src/components/entity_usage.tsx
View file

@ -22,14 +22,10 @@ import {
Subtitle,
} from "@tremor/react";
import UsageDatePicker from "./shared/usage_date_picker";
import { Select } from "antd";
import { ActivityMetrics, processActivityData } from "./activity_metrics";
import {
DailyData,
KeyMetricWithMetadata,
EntityMetricWithMetadata,
} from "./usage/types";
import { tagDailyActivityCall, teamDailyActivityCall } from "./networking";
import { Select } from 'antd';
import { ActivityMetrics, processActivityData } from './activity_metrics';
import { DailyData, BreakdownMetrics, KeyMetricWithMetadata, EntityMetricWithMetadata } from './usage/types';
import { tagDailyActivityCall, teamDailyActivityCall } from './networking';
import TopKeyView from "./top_key_view";
import { formatNumberWithCommas } from "@/utils/dataUtils";
import { valueFormatterSpend } from "./usage/utils/value_formatters";
@ -49,13 +45,6 @@ interface EntityMetrics {
metadata: Record<string, any>;
}
interface BreakdownMetrics {
models: Record<string, any>;
providers: Record<string, any>;
api_keys: Record<string, any>;
entities: Record<string, EntityMetrics>;
}
interface ExtendedDailyData extends DailyData {
breakdown: BreakdownMetrics;
}
@ -203,7 +192,8 @@ const EntityUsage: React.FC<EntityUsageProps> = ({
},
metadata: {
key_alias: metrics.metadata.key_alias,
},
team_id: metrics.metadata.team_id || null
}
};
}
keySpend[key].metrics.spend += metrics.metrics.spend;
@ -297,9 +287,9 @@ const EntityUsage: React.FC<EntityUsageProps> = ({
cache_creation_input_tokens: 0,
},
metadata: {
alias: data.metadata.team_alias || entity,
id: entity,
},
alias: (data.metadata as any).team_alias || entity,
id: entity
}
};
}
entitySpend[entity].metrics.spend += data.metrics.spend;

48
ui/litellm-dashboard/src/components/new_usage.tsx
View file

@ -126,6 +126,7 @@ const NewUsagePage: React.FC<NewUsagePageProps> = ({
cache_creation_input_tokens: 0,
},
metadata: {},
api_key_breakdown: {}
};
}
modelSpend[model].metrics.spend += metrics.metrics.spend;
@ -162,24 +163,24 @@ const NewUsagePage: React.FC<NewUsagePageProps> = ({
// Calculate provider spend from the breakdown data
const getProviderSpend = () => {
const providerSpend: { [key: string]: MetricWithMetadata } = {};
userSpendData.results.forEach((day) => {
Object.entries(day.breakdown.providers || {}).forEach(
([provider, metrics]) => {
if (!providerSpend[provider]) {
providerSpend[provider] = {
metrics: {
spend: 0,
prompt_tokens: 0,
completion_tokens: 0,
total_tokens: 0,
api_requests: 0,
successful_requests: 0,
failed_requests: 0,
cache_read_input_tokens: 0,
cache_creation_input_tokens: 0,
},
metadata: {},
};
userSpendData.results.forEach(day => {
Object.entries(day.breakdown.providers || {}).forEach(([provider, metrics]) => {
if (!providerSpend[provider]) {
providerSpend[provider] = {
metrics: {
spend: 0,
prompt_tokens: 0,
completion_tokens: 0,
total_tokens: 0,
api_requests: 0,
successful_requests: 0,
failed_requests: 0,
cache_read_input_tokens: 0,
cache_creation_input_tokens: 0
},
metadata: {},
api_key_breakdown: {}
};
}
providerSpend[provider].metrics.spend += metrics.metrics.spend;
providerSpend[provider].metrics.prompt_tokens +=
@ -198,8 +199,7 @@ const NewUsagePage: React.FC<NewUsagePageProps> = ({
metrics.metrics.cache_read_input_tokens || 0;
providerSpend[provider].metrics.cache_creation_input_tokens +=
metrics.metrics.cache_creation_input_tokens || 0;
}
);
});
});
return Object.entries(providerSpend).map(([provider, metrics]) => ({
@ -232,7 +232,8 @@ const NewUsagePage: React.FC<NewUsagePageProps> = ({
},
metadata: {
key_alias: metrics.metadata.key_alias,
},
team_id: null
}
};
}
keySpend[key].metrics.spend += metrics.metrics.spend;
@ -331,6 +332,7 @@ const NewUsagePage: React.FC<NewUsagePageProps> = ({
const modelMetrics = processActivityData(userSpendData, "models");
const keyMetrics = processActivityData(userSpendData, "api_keys");
const mcpServerMetrics = processActivityData(userSpendData, "mcp_servers");
return (
<div style={{ width: "100%" }} className="p-8">
@ -380,6 +382,7 @@ const NewUsagePage: React.FC<NewUsagePageProps> = ({
<Tab>Cost</Tab>
<Tab>Model Activity</Tab>
<Tab>Key Activity</Tab>
<Tab>MCP Server Activity</Tab>
</TabList>
<TabPanels>
{/* Cost Panel */}
@ -644,6 +647,9 @@ const NewUsagePage: React.FC<NewUsagePageProps> = ({
<TabPanel>
<ActivityMetrics modelMetrics={keyMetrics} />
</TabPanel>
<TabPanel>
<ActivityMetrics modelMetrics={mcpServerMetrics} />
</TabPanel>
</TabPanels>
</TabGroup>
</TabPanel>

39
ui/litellm-dashboard/src/components/usage/types.ts
View file

@ -18,21 +18,35 @@ export interface DailyData {
export interface BreakdownMetrics {
models: { [key: string]: MetricWithMetadata };
mcp_servers: { [key: string]: MetricWithMetadata };
providers: { [key: string]: MetricWithMetadata };
api_keys: { [key: string]: KeyMetricWithMetadata };
entities: { [key: string]: MetricWithMetadata };
}
export interface MetricWithMetadata {
metrics: SpendMetrics;
metadata: object;
api_key_breakdown: { [key: string]: KeyMetricWithMetadata };
}
export interface KeyMetricWithMetadata {
metrics: SpendMetrics;
metadata: {
key_alias: string | null;
team_id?: string | null;
};
metadata: KeyMetadata;
}
export interface KeyMetadata {
key_alias: string | null;
team_id: string | null;
}
export interface TopApiKeyData {
api_key: string;
key_alias: string | null;
team_id: string | null;
spend: number;
requests: number;
tokens: number;
}
export interface ModelActivityData {
@ -46,6 +60,7 @@ export interface ModelActivityData {
prompt_tokens: number;
completion_tokens: number;
total_spend: number;
top_api_keys: TopApiKeyData[];
daily_data: {
date: string;
metrics: {
@ -62,11 +77,6 @@ export interface ModelActivityData {
}[];
}
export interface KeyMetadata {
key_alias: string | null;
team_id: string | null;
}
export interface EntityMetadata {
alias: string;
id: string;
@ -76,14 +86,3 @@ export interface EntityMetricWithMetadata {
metrics: SpendMetrics;
metadata: EntityMetadata;
}
export interface MetricWithMetadata {
metrics: SpendMetrics;
metadata: object;
}
export interface BreakdownMetrics {
models: { [key: string]: MetricWithMetadata };
providers: { [key: string]: MetricWithMetadata };
api_keys: { [key: string]: KeyMetricWithMetadata };
}