langchain-ai/langchain · error · ValueError

Vectorstore has not implemented the delete method

Error message

Vectorstore has not implemented the delete method

What it means

Raised in `index()` preflight when the destination's `delete` method is still the base `VectorStore.delete` implementation. The base class deliberately makes `delete` raise `NotImplementedError`; the check `type(destination).delete == VectorStore.delete` detects subclasses that never overrode it. Running cleanup against such a store would fail mid-indexing, so it fails fast with a `ValueError` instead.

Source

Thrown at libs/core/langchain_core/indexing/api.py:442

    destination = vector_store  # Renaming internally for clarity

    # If it's a vectorstore, let's check if it has the required methods.
    if isinstance(destination, VectorStore):
        # Check that the Vectorstore has required methods implemented
        methods = ["delete", "add_documents"]

        for method in methods:
            if not hasattr(destination, method):
                msg = (
                    f"Vectorstore {destination} does not have required method {method}"
                )
                raise ValueError(msg)

        if type(destination).delete == VectorStore.delete:
            # Checking if the VectorStore has overridden the default delete method
            # implementation which just raises a NotImplementedError
            msg = "Vectorstore has not implemented the delete method"
            raise ValueError(msg)
    elif isinstance(destination, DocumentIndex):
        pass
    else:
        msg = (  # type: ignore[unreachable]
            f"Vectorstore should be either a VectorStore or a DocumentIndex. "
            f"Got {type(destination)}."
        )
        raise TypeError(msg)

    if isinstance(docs_source, BaseLoader):
        try:
            doc_iterator = docs_source.lazy_load()
        except NotImplementedError:
            doc_iterator = iter(docs_source.load())
    else:
        doc_iterator = iter(docs_source)

    source_id_assigner = _get_source_id_assigner(source_id_key)

View on GitHub (pinned to e32fa9a52e)

Solutions

1. Override `delete` in your VectorStore subclass to call the backend's removal API.
2. If deletion is unsupported by design, subclass such that indexing is not used, or raise your own descriptive error.
3. Check for an updated version of the third-party store package — many added delete support once this check landed.

Example fix

# before
class MyStore(VectorStore):
    def add_documents(self, documents, **kwargs): ...
    # delete inherited (NotImplementedError)

# after
class MyStore(VectorStore):
    def add_documents(self, documents, **kwargs): ...
    def delete(self, ids=None, **kwargs):
        self._collection.delete(ids=ids)
        return True

from langchain_core.vectorstores import VectorStore

if type(store).delete is VectorStore.delete:
    raise TypeError("store has not overridden delete(); indexing cleanup would fail")
index(store, docs, rm, cleanup="incremental", source_id_key="source")

from langchain_core.vectorstores import VectorStore

def has_real_delete(store: VectorStore) -> bool:
    return type(store).delete is not VectorStore.delete

When it happens

Trigger: Subclassing `VectorStore` (or a mid-tier base like `EmbeddingsStore`) and implementing `add_documents` + `similarity_search` but not `delete`, then calling `index(..., cleanup=...)`. Also triggered by cleanup=None? No — this check runs before cleanup branching in this code path when a VectorStore destination is used with this indexer version.

Common situations: Minimal custom stores built by copying only the search methods from an example; older third-party stores predating the `delete` requirement; wrappers that inherit delete unchanged from an intermediate class that also never implemented it.

Related errors

• The delete operation to VectorStore failed.
• Vectorstore {destination} does not have required method {met
• The delete operation to DocumentIndex failed.
• Vectorstore should be either a VectorStore or a DocumentInde
• Vectorstore should be either a VectorStore or a DocumentInde

AI-assisted analysis of langchain-ai/langchain@e32fa9a52e (2026-08-14). Data as JSON: /api/errors/abce3e5f76ed7117. Report an issue: GitHub.