run-llama/llama_index · error · NotImplementedError

This query engine does not support retrieve, use query direc

Error message

This query engine does not support retrieve, use query directly

What it means

BaseQueryEngine implements only query/_query and aquery/_aquery. The retrieve() method exists only for interface parity with retriever-style pipelines and always raises NotImplementedError on the base class — a query engine produces a full response, not a node list, so 'retrieve' is not a meaningful operation for it.

Source

Thrown at llama-index-core/llama_index/core/base/base_query_engine.py:63

        dispatcher.event(
            QueryEndEvent(query=str_or_query_bundle, response=query_result)
        )
        return query_result

    @dispatcher.span
    async def aquery(self, str_or_query_bundle: QueryType) -> RESPONSE_TYPE:
        dispatcher.event(QueryStartEvent(query=str_or_query_bundle))
        with self.callback_manager.as_trace("query"):
            if isinstance(str_or_query_bundle, str):
                str_or_query_bundle = QueryBundle(str_or_query_bundle)
            query_result = await self._aquery(str_or_query_bundle)
        dispatcher.event(
            QueryEndEvent(query=str_or_query_bundle, response=query_result)
        )
        return query_result

    def retrieve(self, query_bundle: QueryBundle) -> List[NodeWithScore]:
        raise NotImplementedError(
            "This query engine does not support retrieve, use query directly"
        )

    def synthesize(
        self,
        query_bundle: QueryBundle,
        nodes: List[NodeWithScore],
        additional_source_nodes: Optional[Sequence[NodeWithScore]] = None,
    ) -> RESPONSE_TYPE:
        raise NotImplementedError(
            "This query engine does not support synthesize, use query directly"
        )

    async def asynthesize(
        self,
        query_bundle: QueryBundle,
        nodes: List[NodeWithScore],
        additional_source_nodes: Optional[Sequence[NodeWithScore]] = None,

View on GitHub (pinned to afd0fef371)

Solutions

1. Call .query(query_bundle) / await .aquery(...) on query engines and use .retrieve() only on BaseRetriever objects.
2. If you need nodes from a query engine, use its underlying retriever (e.g. RetrieverQueryEngine._retriever) or query then convert str(response) to a TextNode — mirroring base_retriever's own handling.
3. Where an IndexNode needs a retrievable object, wrap the engine's retriever instead of the engine.

Example fix

# before
nodes = query_engine.retrieve(QueryBundle("what is this?"))  # NotImplementedError

# after
response = query_engine.query("what is this?")
# or, if nodes are needed, use the underlying retriever:
nodes = retriever.retrieve(QueryBundle("what is this?"))

from llama_index.core.base.retriever import BaseRetriever

def get_nodes(obj, query_bundle):
    if isinstance(obj, BaseRetriever):
        return obj.retrieve(query_bundle)
    return None  # query engines: use query() instead

from llama_index.core.base.query_engine import BaseQueryEngine
from llama_index.core.base.retriever import BaseRetriever

def supports_retrieve(obj) -> bool:
    return isinstance(obj, BaseRetriever)

When it happens

Trigger: Calling query_engine.retrieve(query_bundle) on any standard query engine (RetrieverQueryEngine, etc.); generic code that receives both retrievers and query engines and uniformly calls .retrieve().

Common situations: Plugging a query engine where a retriever is expected (e.g. QueryFusionRetriever retrievers list, recursion via IndexNode expecting a retriever); helper functions typed loosely that assume the retrieve API exists.

Related errors

• This query engine does not support synthesize, use query dir
• This query engine does not support asynthesize, use aquery d
• Not supported
• This query engine does not support _query.
• This query engine does not support _aquery.

AI-assisted analysis of run-llama/llama_index@afd0fef371 (2026-08-15). Data as JSON: /api/errors/e2f10872e064052c. Report an issue: GitHub.