BerriAI/litellm · error · HTTPException

tpm limit higher than user max. User tpm limit={user_api_key

Error message

tpm limit higher than user max. User tpm limit={user_api_key_dict.tpm_limit}. User role={user_api_key_dict.user_role}

What it means

TPM self-limit guard: the caller requests a team tpm_limit above their own user tpm_limit, which would delegate more guaranteed throughput than they hold; refused with 400 naming the user's tpm limit and role.

Source

Thrown at litellm/proxy/management_endpoints/team_endpoints.py:1080

    # Validate team models against user's allowed models
    if data.models is not None and len(user_api_key_dict.models) > 0:
        for m in data.models:
            if m not in user_api_key_dict.models:
                raise HTTPException(
                    status_code=400,
                    detail={
                        "error": f"Model not in allowed user models. User allowed models={user_api_key_dict.models}. User id={user_api_key_dict.user_id}"
                    },
                )

    # Validate team TPM/RPM against user's TPM/RPM limits
    if (
        data.tpm_limit is not None
        and user_api_key_dict.tpm_limit is not None
        and data.tpm_limit > user_api_key_dict.tpm_limit
    ):
        raise HTTPException(
            status_code=400,
            detail={
                "error": f"tpm limit higher than user max. User tpm limit={user_api_key_dict.tpm_limit}. User role={user_api_key_dict.user_role}"
            },
        )

    if (
        data.rpm_limit is not None
        and user_api_key_dict.rpm_limit is not None
        and data.rpm_limit > user_api_key_dict.rpm_limit
    ):
        raise HTTPException(
            status_code=400,
            detail={
                "error": f"rpm limit higher than user max. User rpm limit={user_api_key_dict.rpm_limit}. User role={user_api_key_dict.user_role}"
            },
        )

View on GitHub (pinned to 77b7c6c40c)

Solutions

1. Set tpm_limit at or below your user's tpm limit.

When it happens

Trigger: Thrown at litellm/proxy/management_endpoints/team_endpoints.py:1080 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of BerriAI/litellm@77b7c6c40c (2026-08-18). Data as JSON: /api/errors/a6c4f7d301a5b797. Report an issue: GitHub.