BerriAI/litellm · error · ValueError

Bedrock Mantle auth failed: no Bearer token and no usable AW

Error message

Bedrock Mantle auth failed: no Bearer token and no usable AWS credentials. Set BEDROCK_MANTLE_API_KEY (or AWS_BEARER_TOKEN_BEDROCK) or pass api_key for Bearer auth, or provide AWS credentials (IAM role / access key / profile / web identity) for SigV4.

What it means

Error "Bedrock Mantle auth failed: no Bearer token and no usable AWS credentials. Set BEDROCK_MANTLE_API_KEY (or AWS_BEARER_TOKEN_BEDROCK) or pass api_key for Bearer auth, or provide AWS credentials (IAM role / access key / profile / web identity) for SigV4." thrown in BerriAI/litellm.

Source

Thrown at litellm/llms/bedrock_mantle/common_utils.py:103

        try:
            return self._aws_signer._sign_request(
                service_name="bedrock",
                headers=headers,
                optional_params=optional_params,
                request_data=request_data,
                api_base=api_base,
                api_key=bearer,
                model=model,
                stream=stream,
                fake_stream=fake_stream,
            )
        except (
            NoCredentialsError,
            PartialCredentialsError,
            ProfileNotFound,
            CredentialRetrievalError,
        ) as e:
            raise ValueError(
                "Bedrock Mantle auth failed: no Bearer token and no usable AWS "
                "credentials. Set BEDROCK_MANTLE_API_KEY (or AWS_BEARER_TOKEN_BEDROCK) "
                "or pass api_key for Bearer auth, or provide AWS credentials "
                "(IAM role / access key / profile / web identity) for SigV4."
            ) from e


def mantle_supports_responses(model: str | None, model_cost: dict) -> bool:
    """Whether a Bedrock Mantle model can serve the native Responses API.

    Purely data-driven from the model's price-map capability signal -- either
    /v1/responses in supported_endpoints, or mode=responses -- both overridable
    via register_model and proxy model_info, so onboarding a model is a JSON
    change, never a code change. There is deliberately NO model-name match here:
    capability is per-model, not per-family (openai.gpt-oss-120b supports
    Responses while openai.gpt-oss-safeguard-120b does not, despite sharing the
    gpt-oss substring), so a substring gate would be wrong. A model absent from
    model_cost simply has no signal and returns False (chat-completions emulation).

View on GitHub (pinned to 6c2dcb801b)

Solutions

1. Set BEDROCK_MANTLE_API_KEY (Bearer auth) or provide AWS credentials for SigV4.

When it happens

Trigger: Thrown at litellm/llms/bedrock_mantle/common_utils.py:103 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of BerriAI/litellm@6c2dcb801b (2026-08-15). Data as JSON: /api/errors/edc369b88755dce4. Report an issue: GitHub.