Custom Auth

You can now override the default api key auth.

Enforcement with custom auth

By default, custom auth enforces only the rate limits you set on the returned object. Budgets and model-access require a flag. The table below shows, for each control, where to configure it and which flags it needs.

What gets enforced

Goal	Set it on	Flags required
Key / user / team / end-user rate limits	returned object (rpm_limit, team_tpm_limit, …)	none
Per-model rate limits, key / team scoped	metadata / team_metadata on the returned object	none
Per-model rate limits, project scoped	the project record (model_tpm_limit / model_rpm_limit)	custom_auth_run_common_checks
Team / user / project budget	the team / user / project record	custom_auth_run_common_checks
Team / user / project model allowlist	the team / user / project record	custom_auth_run_common_checks
End-user budget	the end-user record	custom_auth_run_common_checks or enable_post_custom_auth_checks
Key model allowlist (models)	returned object	both flags
Key per-model budget (model_max_budget)	returned object	enable_post_custom_auth_checks
Key expiry (expires)	returned object	enable_post_custom_auth_checks
Key scalar budget (max_budget / soft_budget)	not supported; use a per-scope budget	n/a

Note: Project per-model limits go on the object's project_metadata with the flag off, but the DB project record overrides it once the flag is on, so set them there. (Team per-model via team_metadata always stays on the object.)

See Enforce budgets and model access and Key-level enforcement for examples.

Usage

1. Create a custom auth file.

Make sure the response type follows the UserAPIKeyAuth pydantic object. This is used by for logging usage specific to that user key.

from fastapi import Request
from litellm.proxy._types import UserAPIKeyAuth

async def user_api_key_auth(request: Request, api_key: str) -> UserAPIKeyAuth: 
    try: 
        modified_master_key = "sk-my-master-key"
        if api_key == modified_master_key:
            return UserAPIKeyAuth(api_key=api_key)
        raise Exception
    except: 
        raise Exception

2. Pass the filepath (relative to the config.yaml)

Pass the filepath to the config.yaml

e.g. if they're both in the same dir - ./config.yaml and ./custom_auth.py, this is what it looks like:

model_list: 
  - model_name: "openai-model"
    litellm_params: 
      model: "gpt-3.5-turbo"

litellm_settings:
  drop_params: True
  set_verbose: True

general_settings:
  custom_auth: custom_auth.user_api_key_auth

Implementation Code

3. Start the proxy

$ litellm --config /path/to/config.yaml 

UserAPIKeyAuth Fields Reference

These fields are read straight off the returned object and enforced with no flag. Budgets and model-access are enforced behind flags (see below).

Identity

Who the request belongs to. The *_id fields also tell LiteLLM which DB records to load when custom_auth_run_common_checks: true.

UserAPIKeyAuth(
    api_key: Optional[str] = None,                    # The API key (will be hashed automatically)
    token: Optional[str] = None,                      # Hashed token for internal use
    key_alias: Optional[str] = None,                  # Key alias for identification
    user_id: Optional[str] = None,                    # User identifier (also used to load the user record)
    user_email: Optional[str] = None,                 # User email address
    user_role: Optional[LitellmUserRoles] = None,     # User role (PROXY_ADMIN, INTERNAL_USER, etc.)
    team_id: Optional[str] = None,                    # Team identifier (also used to load the team record)
    org_id: Optional[str] = None,                     # Organization identifier (also used to load the org record)
    end_user_id: Optional[str] = None,                # End-user identifier (also used to load the end-user record)
)

Rate limits

All scopes below are enforced directly off the returned object, with no flag.

UserAPIKeyAuth(
    # Key
    tpm_limit: Optional[int] = None,
    rpm_limit: Optional[int] = None,
    # User
    user_tpm_limit: Optional[int] = None,
    user_rpm_limit: Optional[int] = None,
    # Team
    team_tpm_limit: Optional[int] = None,
    team_rpm_limit: Optional[int] = None,
    # Per team-member
    team_member_tpm_limit: Optional[int] = None,
    team_member_rpm_limit: Optional[int] = None,
    # Per end-user
    end_user_tpm_limit: Optional[int] = None,
    end_user_rpm_limit: Optional[int] = None,
    # Per-model (key / team scoped)
    metadata: Dict = {},          # e.g. {"model_tpm_limit": {...}, "model_rpm_limit": {...}}
    team_metadata: Optional[Dict] = None,  # same keys, team scoped
)

備註

Per-model rate limits are read from metadata (key) and team_metadata (team), keyed by model name. The model key must equal the request's model string exactly, or the limit is skipped silently.

rpm_limit_per_model / tpm_limit_per_model exist on the object but are inert; use metadata / team_metadata instead, or the project record (see below).

Advanced

UserAPIKeyAuth(
    max_parallel_requests: Optional[int] = None,      # Concurrent request limit
    allowed_model_region: Optional[AllowedModelRegion] = None,  # Geographic restrictions
    blocked: Optional[bool] = None,                   # Whether the key is blocked
    config: Dict = {},                                # Configuration settings
)

Object permission (MCP, agents, etc.)

from litellm.proxy._experimental.mcp_server.mcp_server_manager import (
    global_mcp_server_manager,
)

def _server_id(name: str) -> str:
    server = global_mcp_server_manager.get_mcp_server_by_name(name)
    if not server:
        raise ValueError(f"Unknown MCP server '{name}'")
    return server.server_id

object_permission = LiteLLM_ObjectPermissionTable(
    mcp_servers=[_server_id("deepwiki"), _server_id("everything")], # MCP servers this key is allowed to use
    mcp_tool_permissions={"deepwiki": ["search", "read_doc"]},      # optional per-server tool allow-list
)

UserAPIKeyAuth(
    object_permission=object_permission,
)

Enforce budgets and model access

Set custom_auth_run_common_checks: true to enforce budgets and model-access alongside custom auth:

general_settings:
  custom_auth: custom_auth.user_api_key_auth
  custom_auth_run_common_checks: true

Your handler returns the IDs; the budgets and allowlists live on the matching DB records (/team/new, /user/new, /project/new, /customer/new, or the UI), which LiteLLM loads and enforces.

For example, a team with a budget and model allowlist:

curl -X POST 'http://0.0.0.0:4000/team/new' \
  -H 'Authorization: Bearer sk-master-key' \
  -H 'Content-Type: application/json' \
  -d '{
    "team_id": "eng-team",
    "max_budget": 100,
    "models": ["gpt-4o-mini", "claude-3-haiku"]
  }'

# ...then return that team_id from custom auth:
return UserAPIKeyAuth(api_key=api_key, team_id="eng-team")

For project per-model rate limits, set model_tpm_limit / model_rpm_limit on the project record (keyed by model name) and return that project_id:

# On the project record (via /project/new or the UI):
#   model_tpm_limit = {"gpt-4o": 100000, "claude-3-haiku": 50000}
#   model_rpm_limit = {"gpt-4o": 100,    "claude-3-haiku": 200}

備註

• The project record's metadata replaces any project_metadata you set on the returned object, so configure project per-model limits on the project record, not on the object.
• For per-model rate limits, the model key must equal the request's model string exactly, or the limit is skipped silently. This was the actual Expedia failure mode.

Key models vs project models

These are separate controls:

Field	Where it is enforced	Source of truth
models on UserAPIKeyAuth	Key-level allowlist	Value you return from custom auth
project_id on UserAPIKeyAuth	Project-level allowlist	models on the project record in LiteLLM's DB

An empty models list ([]) means no restriction. Names must match the model group in your config (wildcards supported). See Project Management and Config Settings.

Key-level enforcement

The following are enforced from the returned object, but only when litellm.enable_post_custom_auth_checks: true is also set:

general_settings:
  custom_auth: custom_auth.user_api_key_auth
  custom_auth_run_common_checks: true   # required for the key models allowlist

litellm_settings:
  enable_post_custom_auth_checks: true

from datetime import datetime, timedelta, timezone

return UserAPIKeyAuth(
    api_key=api_key,
    models=["gpt-4o-mini"],                                   # key model allowlist (needs both flags)
    model_max_budget={"gpt-4o": {"budget_limit": 100, "time_period": "30d"}},  # key per-model budget
    expires=datetime.now(timezone.utc) + timedelta(days=30),  # key expiry
)

This path also enforces end-user budgets and per-model end-user budgets when end_user_id is set.

✨ Support LiteLLM Virtual Keys + Custom Auth

Supported from v1.72.2+

資訊

✨ Supporting Custom Auth + LiteLLM Virtual Keys is on LiteLLM Enterprise

Enterprise Pricing

Get free 7-day trial key

Usage

1. Setup custom auth file

"""
Example custom auth function.

This will allow all keys starting with "my-custom-key" to pass through.
"""
from typing import Union

from fastapi import Request

from litellm.proxy._types import UserAPIKeyAuth


async def user_api_key_auth(
    request: Request, api_key: str
) -> Union[UserAPIKeyAuth, str]:
    try:
        if api_key.startswith("my-custom-key"):
            return "sk-P1zJMdsqCPNN54alZd_ETw"
        else:
            raise Exception("Invalid API key")
    except Exception:
        raise Exception("Invalid API key")

2. Setup config.yaml

Key change set mode: auto. This will check both litellm api key auth + custom auth.

model_list: 
  - model_name: "openai-model"
    litellm_params: 
      model: "gpt-3.5-turbo"
      api_key: os.environ/OPENAI_API_KEY

general_settings:
  custom_auth: custom_auth_auto.user_api_key_auth
  custom_auth_settings:
    mode: "auto" # can be 'on', 'off', 'auto' - 'auto' checks both litellm api key auth + custom auth

Flow:

1. Checks custom auth first

2. If custom auth fails, checks litellm api key auth

3. If both fail, returns 401

4. Test it!

curl -L -X POST 'http://0.0.0.0:4000/v1/chat/completions' \
-H 'Content-Type: application/json' \
-H 'Authorization: Bearer sk-P1zJMdsqCPNN54alZd_ETw' \
-d '{
    "model": "openai-model",
    "messages": [
          {
            "role": "user",
            "content": "Hey! My name is John"
          }
        ]
}'

Bubble up custom exceptions

If you want to bubble up custom exceptions, you can do so by raising a ProxyException.

"""
Example custom auth function.

This will allow all keys starting with "my-custom-key" to pass through.
"""

from typing import Union

from fastapi import Request

from litellm.proxy._types import UserAPIKeyAuth, ProxyException


async def user_api_key_auth(
    request: Request, api_key: str
) -> Union[UserAPIKeyAuth, str]:
    try:
        if api_key.startswith("my-custom-key"):
            return "sk-P1zJMdsqCPNN54alZd_ETw"
        if api_key == "invalid-api-key":
            # raise a custom exception back to the client
            raise ProxyException(
                message="Invalid API key",
                type="invalid_request_error",
                param="api_key",
                code=401,
            )
        else:
            raise Exception("Invalid API key")
    except Exception:
        raise Exception("Invalid API key")