Comfy-Org/ComfyUI · error · ValueError

Gemini interaction did not complete (status: {interaction.st

Error message

Gemini interaction did not complete (status: {interaction.status}).

What it means

After a Gemini Interactions API call returns, the node requires interaction.status == 'completed'. Any other status (e.g. 'failed', 'cancelled', or an in-progress state) raises this error, appending the model's text output when present so the failure reason is visible. This is a post-response check — the request itself succeeded at the transport level but the interaction did not finish generating.

Source

Thrown at comfy_api_nodes/nodes_gemini.py:1689

            )
            parts.extend(to_interaction_media_part(p) for p in media_parts)
        parts.append(GeminiInteractionTextPart(text=prompt))
        interaction = await sync_op(
            cls,
            ApiEndpoint(path=GEMINI_INTERACTIONS_ENDPOINT, method="POST"),
            data=GeminiInteractionRequest(
                model=model_id,
                input=parts,
                generation_config=GeminiInteractionGenerationConfig(
                    temperature=model.get("temperature", 1.0),
                    top_p=model.get("top_p", 0.95),
                ),
            ),
            response_model=GeminiInteraction,
        )
        if interaction.status != "completed":
            model_message = get_text_from_interaction(interaction).strip()
            raise ValueError(
                f"Gemini interaction did not complete (status: {interaction.status})."
                + (f" Model response: {model_message}" if model_message else "")
            )
        return IO.NodeOutput(
            await get_video_from_interaction(interaction, cls=cls),
            get_text_from_interaction(interaction),
        )


class GeminiExtension(ComfyExtension):
    @override
    async def get_node_list(self) -> list[type[IO.ComfyNode]]:
        return [
            GeminiNode,
            GeminiNodeV2,
            GeminiImage,
            GeminiImage2,
            GeminiNanoBanana2,

View on GitHub (pinned to 1c6d8d45b3)

Solutions

Read the appended 'Model response:' text, when present, for the concrete failure reason.
Retry the interaction — non-completed statuses are often transient.
Reduce generation complexity: shorter duration, fewer inputs, simpler prompt.
If persistent, check the Gemini API status page / try a different omni model id.

Example fix

// before
interaction = await gemini_interactions_generate(...)
video = await get_video_from_interaction(interaction, cls=cls)  # raises: did not complete (status: failed)

// after
interaction = await gemini_interactions_generate(...)
if interaction.status != 'completed':
    raise RuntimeError(f"interaction {interaction.status}: {get_text_from_interaction(interaction).strip()}")
video = await get_video_from_interaction(interaction, cls=cls)

Defensive patterns

Strategy: retry

Try / catch

try:
    interaction = await gemini_interactions_generate(...)
    assert interaction.status == "completed", get_text_from_interaction(interaction).strip()
except (ValueError, AssertionError) as e:
    if is_transient(e):  # failed/cancelled without model explanation
        interaction = await gemini_interactions_generate(...)  # one retry
    else:
        raise

Prevention

Keep inline payloads comfortably below the 90 MiB cap to avoid server-side aborts.
Prefer shorter durations / fewer inputs for the first run of a new workflow.
Surface interaction.status and the model text together in logs for fast diagnosis.

When it happens

Trigger: gemini_interactions_generate returns a GeminiInteraction whose status field is not 'completed' — provider-side generation failure, safety stop mid-interaction, timeout/cancellation of a long video generation, or an API change introducing a new status value.

Common situations: Long video generations interrupted server-side; content policy stops that abort the interaction rather than returning a refusal message; transient infra failures during preview model rollout; very large inline payloads (near the 90 MiB cap) causing server-side aborts.

Related errors

AI-assisted analysis of Comfy-Org/ComfyUI@1c6d8d45b3 (2026-08-14). Data as JSON: /api/errors/6568a7540d827998. Report an issue: GitHub.