langchain-ai/langchain · error · NotImplementedError

Unsupported message type: {type(message)}

Error message

Unsupported message type: {type(message)}

What it means

NotImplementedError raised in _convert_to_message_template when a message passed to from_messages is not one of the supported types: Base(BaseChat)Message(Base)PromptTemplate, BaseMessage, str, tuple, or dict. This is the catch-all for unsupported input shapes.

Source

Thrown at libs/core/langchain_core/prompts/chat.py:1489

                message_type_str, template, template_format=template_format
            )
        elif (
            hasattr(message_type_str, "model_fields")
            and "type" in message_type_str.model_fields
        ):
            message_type = message_type_str.model_fields["type"].default
            message_ = _create_template_from_message_type(
                message_type, template, template_format=template_format
            )
        else:
            message_ = message_type_str(
                prompt=PromptTemplate.from_template(
                    cast("str", template), template_format=template_format
                )
            )
    else:
        msg = f"Unsupported message type: {type(message)}"  # type: ignore[unreachable]
        raise NotImplementedError(msg)

    return message_


# For backwards compat:
_convert_to_message = _convert_to_message_template

View on GitHub (pinned to e32fa9a52e)

Solutions

1. Check `type(message)` in the error output and convert the item to str/tuple/dict/BaseMessage.
2. Flatten nested lists: from_messages([*inner]) instead of from_messages([inner]).
3. Guard dynamic pipelines by asserting each item is str/tuple/dict/BaseMessage before building the prompt.

Example fix

# before
ChatPromptTemplate.from_messages([[("human", "hi")]])

# after
ChatPromptTemplate.from_messages([("human", "hi")])

from langchain_core.messages import BaseMessage

def supported_message(m):
    return isinstance(m, (str, tuple, dict, BaseMessage))

from langchain_core.messages import BaseMessage
from typing import Any

def is_message_like(m: Any) -> bool:
    return isinstance(m, (str, tuple, dict, BaseMessage))

try:
    prompt = ChatPromptTemplate.from_messages(messages)
except NotImplementedError as e:
    raise ValueError(f"Unsupported message shape in list: {e}") from e

When it happens

Trigger: `from_messages([42])`, `from_messages([None])`, `from_messages([b"bytes message"])`, or passing a generator/iterable of messages as a single element rather than unpacking it.

Common situations: Passing bytes, None, or a number because a variable was never rendered; passing a nested list [[...]] instead of splatting; custom objects that aren't message subclasses.

Related errors

• Invalid placeholder template: {template}. Expected a variabl
• Unexpected arguments for placeholder message type. Expected
• Expected is_optional to be a boolean. Got: {is_optional}
• Expected variable name to be a string. Got: {var_name_wrappe
• Invalid placeholder template: {var_name_wrapped}. Expected a

AI-assisted analysis of langchain-ai/langchain@e32fa9a52e (2026-08-14). Data as JSON: /api/errors/97dd451e514dcc00. Report an issue: GitHub.