invoke-ai/InvokeAI · error · NodeFieldNotFoundError

Edge destination field {edge.destination.field} does not exi

Error message

Edge destination field {edge.destination.field} does not exist in node {edge.destination.node_id}

What it means

NodeFieldNotFoundError is raised when an edge's `destination.field` is not an input field (Pydantic model_fields) of the destination node's type. Dynamic inputs on CallSavedWorkflowInvocation nodes are exempted. The library throws this because writing into a nonexistent input would be silently ignored at execution time.

Source

Thrown at invokeai/app/services/shared/graph.py:1826

            source_node = self.nodes.get(edge.source.node_id, None)
            if source_node is None:
                raise NodeNotFoundError(f"Edge source node {edge.source.node_id} does not exist in the graph")

            destination_node = self.nodes.get(edge.destination.node_id, None)
            if destination_node is None:
                raise NodeNotFoundError(f"Edge destination node {edge.destination.node_id} does not exist in the graph")

            if edge.source.field not in source_node.get_output_annotation().model_fields:
                raise NodeFieldNotFoundError(
                    f"Edge source field {edge.source.field} does not exist in node {edge.source.node_id}"
                )

            if edge.destination.field not in type(destination_node).model_fields:
                if isinstance(destination_node, CallSavedWorkflowInvocation) and is_call_saved_workflow_dynamic_input(
                    edge.destination.field
                ):
                    continue
                raise NodeFieldNotFoundError(
                    f"Edge destination field {edge.destination.field} does not exist in node {edge.destination.node_id}"
                )

    def _validate_graph_is_acyclic(self) -> None:
        graph = self.nx_graph_flat()
        if not nx.is_directed_acyclic_graph(graph):
            raise CyclicalGraphError("Graph contains cycles")

    def _validate_edge_type_compatibility(self) -> None:
        for edge in self.edges:
            destination_node = self.get_node(edge.destination.node_id)
            if isinstance(destination_node, CallSavedWorkflowInvocation) and is_call_saved_workflow_dynamic_input(
                edge.destination.field
            ):
                continue
            if not are_connections_compatible(
                self.get_node(edge.source.node_id),
                edge.source.field,

View on GitHub (pinned to 0b6a024f2f)

Solutions

Change the edge's destination.field to a real input field of the destination node (check type(destination_node).model_fields)
Point the edge at a node type that has the desired input field
Recreate the workflow in the current UI after upgrading InvokeAI
During workflow migration, remap obsolete input names before validating the graph

Example fix

# before
EdgeConnection(node_id="denoise", field="prompt_text")

# after
EdgeConnection(node_id="denoise", field="prompt")  # actual input field

Defensive patterns

Strategy: validation

Validate before calling

def check_edge_destination_fields(graph):
    for edge in graph.edges:
        dst = graph.nodes.get(edge.destination.node_id)
        if dst and edge.destination.field not in type(dst).model_fields:
            raise ValueError(f"{edge.destination.node_id} has no input field {edge.destination.field}")
    return True

check_edge_destination_fields(graph)

Type guard

def is_valid_destination_field(dst, field: str) -> bool:
    return field in type(dst).model_fields or (
        isinstance(dst, CallSavedWorkflowInvocation)
        and is_call_saved_workflow_dynamic_input(field)
    )

Try / catch

from invokeai.app.services.shared.graph import NodeFieldNotFoundError
try:
    validate_graph(graph)
except NodeFieldNotFoundError as e:
    logger.error("edge writes to nonexistent input: %s", e)

Prevention

Validate destination field names against type(node).model_fields before wiring edges
After InvokeAI upgrades, migrate saved workflows and remap renamed inputs
For dynamic-INPUT workflows, only rely on dynamic fields on CallSavedWorkflowInvocation nodes

When it happens

Trigger: Creating an edge whose destination.field does not match any input field of the destination node type — typos, referencing an input that was renamed/removed in a newer InvokeAI version, or feeding a dynamic-INPUT field on a non-CallSavedWorkflowInvocation node.

Common situations: Saved workflows from older InvokeAI versions whose node inputs changed after an upgrade; hand-edited workflow JSON; programmatic graph construction using guessed field names.

Related errors

AI-assisted analysis of invoke-ai/InvokeAI@0b6a024f2f (2026-08-29). Data as JSON: /api/errors/8313f9a7ea1489b1. Report an issue: GitHub.