langchain-ai/langgraph · error · ValueError

embed must be provided

Error message

embed must be provided

What it means

Argument guard at the top of ensure_embeddings: the embed parameter is None. The function's contract is to normalize an Embeddings object, a plain embedding function, or a provider:identifier string into an Embeddings instance; with None there is nothing to wrap, so construction aborts immediately.

Solutions

1. Pass an Embeddings instance, e.g. ensure_embeddings(OpenAIEmbeddings())
2. Pass a callable that maps list[str] -> list[list[float]] to have it wrapped in EmbeddingsLambda
3. Pass a provider string like 'openai:text-embedding-3-small' (requires langchain>=0.3.9)

When it happens

Trigger: Thrown at libs/checkpoint/langgraph/checkpoint/store/base/embed.py:82 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of langchain-ai/langgraph@38031739e5 (2026-08-26). Data as JSON: /api/errors/2e4e4ae37af291c7. Report an issue: GitHub.

Appendix: source

Thrown at libs/checkpoint/langgraph/checkpoint/store/base/embed.py:82

    if embed is None:
        raise ValueError("embed must be provided")
    if isinstance(embed, str):
        init_embeddings = _get_init_embeddings()
        if init_embeddings is None:
            from importlib.metadata import PackageNotFoundError, version

            try:
                lc_version = version("langchain")
                version_info = f"Found langchain version {lc_version}, but"
            except PackageNotFoundError:
                version_info = "langchain is not installed;"

View on GitHub (pinned to 38031739e5)