invoke-ai/InvokeAI · error · ValueError

The VAE must resolve to a standalone Wan VAE model.

Error message

The VAE must resolve to a standalone Wan VAE model.

What it means

_validate_standalone_vae checks that an explicitly provided VAE resolves to a config with base Wan and type VAE. If the field points at any other model kind, invoke() raises this ValueError before channel-count checks. It ensures the standalone VAE override is actually a Wan VAE model.

Source

Thrown at invokeai/app/invocations/wan_model_loader.py:346

    ) -> None:
        WanModelLoaderInvocation._validate_component_source_format(context, model)
        source_config = context.models.get_config(model)
        source_variant = getattr(source_config, "variant", None)
        main_is_ti2v = main_variant == WanVariantType.TI2V_5B
        source_is_ti2v = source_variant == WanVariantType.TI2V_5B
        if main_is_ti2v != source_is_ti2v:
            raise ValueError(
                "The Component Source VAE is incompatible with the selected transformer. "
                "TI2V-5B requires the 48-channel Wan 2.2 VAE; A14B models require the 16-channel Wan 2.1 VAE."
            )

    @staticmethod
    def _validate_standalone_vae(
        context: InvocationContext, model: ModelIdentifierField, main_variant: WanVariantType
    ) -> None:
        vae_config = context.models.get_config(model)
        if vae_config.base != BaseModelType.Wan or vae_config.type != ModelType.VAE:
            raise ValueError("The VAE must resolve to a standalone Wan VAE model.")
        expected_channels = 48 if main_variant == WanVariantType.TI2V_5B else 16
        if vae_config.latent_channels != expected_channels:
            raise ValueError(
                "The standalone VAE is incompatible with the selected transformer. "
                "TI2V-5B requires the 48-channel Wan 2.2 VAE; A14B models require the 16-channel Wan 2.1 VAE."
            )

View on GitHub (pinned to 0b6a024f2f)

Solutions

1. Set 'VAE' to a standalone Wan VAE model (base=Wan, type=VAE)
2. Re-import the VAE if its registered type/base metadata is wrong
3. Clear the override and let the loader use a Diffusers main model or Component Source for the VAE

Example fix

// before
vae=ModelIdentifierField(key='sdxl-vae', submodel_type=SubModelType.VAE)
// after
vae=ModelIdentifierField(key='wan-vae', submodel_type=SubModelType.VAE)

if invocation.vae_model is not None:
    cfg = context.models.get_config(invocation.vae_model)
    if cfg.base != BaseModelType.Wan or cfg.type != ModelType.VAE:
        raise ValueError('VAE field must point at a standalone Wan VAE')

def is_standalone_wan_vae(field, ctx) -> bool:
    cfg = ctx.models.get_config(field)
    return cfg.base == BaseModelType.Wan and cfg.type == ModelType.VAE

try:
    output = invocation.invoke(context)
except ValueError as e:
    if 'must resolve to a standalone Wan VAE model' in str(e):
        invocation.vae_model = select_wan_vae()

When it happens

Trigger: invoke() with the 'VAE' field set to a model whose config.base != Wan or config.type != VAE — e.g. a SDXL VAE, a Wan Main model, or a T5 encoder placed in the VAE field.

Common situations: Selecting the wrong model in the VAE dropdown; reusing a workflow field that previously held a different model type; a VAE imported with incorrect type metadata.

Related errors

• No source for VAE. Either set 'VAE' to a standalone Wan VAE,
• The Wan T5 Encoder must resolve to a standalone Wan T5 encod
• The {label} model must resolve to a Wan main model.
• The Component Source model must resolve to a Wan main model.
• The Component Source VAE is incompatible with the selected t

AI-assisted analysis of invoke-ai/InvokeAI@0b6a024f2f (2026-08-29). Data as JSON: /api/errors/deb5f7359b148e21. Report an issue: GitHub.