langchain-ai/langchain · error · ValueError

Either data or path must be provided

Error message

Either data or path must be provided

What it means

Raised by the `Blob` model validator `check_blob_is_valid` in `langchain_core.documents.base`. A `Blob` represents a piece of content that must be located somewhere: either in memory (`data`) or on the filesystem (`path`). Instantiating a `Blob` with neither field leaves the object with no content to read, so Pydantic rejects it at validation time with a `ValueError`.

Source

Thrown at libs/core/langchain_core/documents/base.py:155

    def source(self) -> str | None:
        """The source location of the blob as string if known otherwise none.

        If a path is associated with the `Blob`, it will default to the path location.

        Unless explicitly set via a metadata field called `'source'`, in which
        case that value will be used instead.
        """
        if self.metadata and "source" in self.metadata:
            return cast("str | None", self.metadata["source"])
        return str(self.path) if self.path else None

    @model_validator(mode="before")
    @classmethod
    def check_blob_is_valid(cls, values: dict[str, Any]) -> Any:
        """Verify that either data or path is provided."""
        if "data" not in values and "path" not in values:
            msg = "Either data or path must be provided"
            raise ValueError(msg)
        return values

    def as_string(self) -> str:
        """Read data as a string.

        Raises:
            ValueError: If the blob cannot be represented as a string.

        Returns:
            The data as a string.
        """
        if self.data is None and self.path:
            return Path(self.path).read_text(encoding=self.encoding)
        if isinstance(self.data, bytes):
            return self.data.decode(self.encoding)
        if isinstance(self.data, str):
            return self.data
        msg = f"Unable to get string for blob {self}"

View on GitHub (pinned to e32fa9a52e)

Solutions

  1. Pass `data` (str or bytes) or `path` (str/Path) when constructing the Blob: `Blob(data="text")` or `Blob.from_path("file.txt")`.
  2. If the value may be None at runtime, guard before construction: `if content is not None: Blob(data=content) else: Blob.from_path(p)`.
  3. Prefer the `Blob.from_path()` / `Blob.from_data()` classmethods, which force you to supply a source explicitly.

Example fix

// before
blob = Blob(metadata={"source": "a.pdf"})

// after
blob = Blob.from_path("a.pdf", metadata={"source": "a.pdf"})
Defensive patterns

Strategy: validation

Validate before calling

def make_blob(*, data=None, path=None, **kw):
    if data is None and path is None:
        raise ValueError("Refusing to create Blob without data or path")
    return Blob(data=data, path=path, **kw)

Type guard

from langchain_core.documents import Blob

def is_valid_blob_source(data, path) -> bool:
    return data is not None or path is not None

Prevention

When it happens

Trigger: Constructing `Blob()` with no arguments, or passing only `metadata`/`mime_type`/`encoding` without `data` or `path`. Also happens when code builds a Blob from a variable that is unexpectedly `None`, e.g. `Blob(data=maybe_none)` — note the key must be present in the values dict, so `Blob(data=None)` actually passes this check but fails later; the error fires only when BOTH keys are absent.

Common situations: Migrating loaders from the legacy `Document(loader, blob=...)` pattern where content was attached elsewhere; building Blobs dynamically from user input where the source may be empty; forgetting to pass through the `data` kwarg in a custom loader's super() call.

Related errors

AI-assisted analysis of langchain-ai/langchain@e32fa9a52e (2026-08-14). Data as JSON: /api/errors/d7a8f9bdd3c34e31. Report an issue: GitHub.