BerriAI/litellm · error · Exception

Setting tag based guardrail modes is only available in litel

Error message

Setting tag based guardrail modes is only available in litellm-enterprise. You must be a LiteLLM Enterprise user to use this feature. If you have a license please set `LITELLM_LICENSE` in your env. Get a 7 day trial key here: https://www.litellm.ai/enterprise#trial. 
Pricing: https://www.litellm.ai/#pricing.

What it means

Raised by custom_guardrail.py in LiteLLM Enterprise when tag-based guardrail modes are configured but the running proxy is not a premium (licensed) user. The hook imports premium_user from litellm.proxy.proxy_server; if the enterprise license check failed, any attempt to resolve a Mode-tagged guardrail raises this exception.

Source

Thrown at enterprise/litellm_enterprise/integrations/custom_guardrail.py:30

        ],
        event_type: Optional[GuardrailEventHooks] = None,
    ) -> Optional[bool]:
        """
        Returns True if the guardrail should be run for this request and event_type.

        Logic:
        - If a request tag matches a Mode tag key, only run if event_type matches
          the tag's value (the mode for that tag).
        - If no request tag matches, fall back to default mode(s).
        """
        from litellm.litellm_core_utils.litellm_logging import (
            StandardLoggingPayloadSetup,
        )
        from litellm.proxy._types import CommonProxyErrors
        from litellm.proxy.proxy_server import premium_user

        if not premium_user:
            raise Exception(
                f"Setting tag based guardrail modes is only available in litellm-enterprise. {CommonProxyErrors.not_premium_user.value}."
            )

        if event_hook is None or not isinstance(event_hook, Mode):
            return None

        proxy_server_request = data.get("proxy_server_request", {})

        request_tags = StandardLoggingPayloadSetup._get_request_tags(
            litellm_params=data,
            proxy_server_request=proxy_server_request,
        )

        # Check if any request tag matches a Mode tag key
        matched_mode = None
        if request_tags:
            for tag in request_tags:
                if tag in event_hook.tags:

View on GitHub (pinned to 6c2dcb801b)

Solutions

1. Set LITELLM_LICENSE to a valid enterprise license key in the proxy environment and restart (get a trial key at litellm.ai/enterprise#trial if evaluating)
2. If you do not have a license, remove tag-based mode configuration from guardrails and use default_mode only
3. Verify premium status at startup: the proxy logs a premium/enterprise banner; check for license validation errors in logs

Example fix

# before
guardrails:
  - guardrail_name: bedrock
    litellm_params:
      guardrail: bedrock
      mode: pre_call_check
      tag_based_modes:
        my_tag: pre_call_check
# no LITELLM_LICENSE set -> error

# after
export LITELLM_LICENSE=<valid-enterprise-key>
# restart proxy

def has_enterprise(proxy_url: str, key: str) -> bool:
    # enterprise features advertise themselves; a cheap proxy is to check config side
    return bool(os.environ.get('LITELLM_LICENSE'))
# or catch deterministically on first request in a canary call

try:
    await litellm.acompletion(**params)  # with tag-based guardrail modes
except Exception as e:
    if 'not_premium_user' in str(e) or 'litellm-enterprise' in str(e):
        raise ConfigError('Set LITELLM_LICENSE or remove tag_based_modes') from e
    raise

When it happens

Trigger: Configuring a guardrail with per-tag modes (e.g. litellm_params tag mappings whose value is a guardrail Mode like 'pre_call_check') and sending a request through a proxy without a valid LITELLM_LICENSE. The mode-resolution path executes on every relevant request once tag-based modes are configured.

Common situations: Trialing tag-based guardrails on an OSS install; LITELLM_LICENSE set but expired or invalid so premium_user resolved False; license env var not passed into the Docker container.

Related errors

• You must be a LiteLLM Enterprise user to use this feature. I
• 🚨🚨🚨 DISABLING LLM API ENDPOINTS is an Enterprise feature
• 🚨🚨🚨 DISABLING ADMIN ENDPOINTS is an Enterprise feature 🚨
• Database not connected
• DB not connected. This endpoint needs a database; set DATABA

AI-assisted analysis of BerriAI/litellm@6c2dcb801b (2026-08-15). Data as JSON: /api/errors/aa4e46b47e1a3c64. Report an issue: GitHub.