invoke-ai/InvokeAI · error · TypeError

Expected AutoencoderKLWan or FluxAutoEncoder for Anima VAE,

Error message

Expected AutoencoderKLWan or FluxAutoEncoder for Anima VAE, got {type(vae_info.model).__name__}.

What it means

The Anima VAE encode entry point only supports AutoencoderKLWan or FluxAutoEncoder VAE models. Before doing any work, vae_encode type-checks the loaded model and raises this TypeError if the model manager returned some other VAE class.

Source

Thrown at invokeai/app/invocations/anima_image_to_latents.py:61

@invocation(
    "anima_i2l",
    title="Image to Latents - Anima",
    tags=["image", "latents", "vae", "i2l", "anima"],
    category="image",
    version="1.0.1",
    classification=Classification.Prototype,
)
class AnimaImageToLatentsInvocation(BaseInvocation, WithMetadata, WithBoard):
    """Generates latents from an image using the Anima VAE (supports Wan 2.1 and FLUX VAE)."""

    image: ImageField = InputField(description="The image to encode.")
    vae: VAEField = InputField(description=FieldDescriptions.vae, input=Input.Connection)

    @staticmethod
    def vae_encode(vae_info: LoadedModel, image_tensor: torch.Tensor) -> torch.Tensor:
        if not isinstance(vae_info.model, (AutoencoderKLWan, FluxAutoEncoder)):
            raise TypeError(
                f"Expected AutoencoderKLWan or FluxAutoEncoder for Anima VAE, got {type(vae_info.model).__name__}."
            )

        if isinstance(vae_info.model, AutoencoderKLWan):
            estimated_working_memory = estimate_vae_working_memory_anima(
                operation="encode",
                image_tensor=image_tensor,
                vae=vae_info.model,
                tile_size=None,
            )
        else:
            estimated_working_memory = estimate_vae_working_memory_flux(
                operation="encode",
                image_tensor=image_tensor,
                vae=vae_info.model,
            )

        with vae_info.model_on_device(working_mem_bytes=estimated_working_memory) as (_, vae):

View on GitHub (pinned to 0b6a024f2f)

Solutions

1. Select an Anima-compatible VAE (AutoencoderKLWan or FluxAutoEncoder) in the VAE loader node.
2. Verify the VAE model's base/architecture matches Anima in the model manager.
3. Re-import the VAE if the model manager resolved the wrong class for its key.

Example fix

// before
vae = vae_loader(vae_model="sdxl-vae")
// after
vae = vae_loader(vae_model="anima-vae")  # AutoencoderKLWan / FluxAutoEncoder

vae_info = context.models.load(vae.vae)
from diffusers import AutoencoderKLWan
from invokeai.backend.flux.model import FluxAutoEncoder
assert isinstance(vae_info.model, (AutoencoderKLWan, FluxAutoEncoder)), "incompatible VAE"

def is_anima_vae(model) -> bool:
    from diffusers import AutoencoderKLWan
    from invokeai.backend.flux.model import FluxAutoEncoder
    return isinstance(model, (AutoencoderKLWan, FluxAutoEncoder))

try:
    latents = img2latents.invoke(context)
except TypeError as e:
    if "Anima VAE" in str(e):
        raise RuntimeError("Attach an Anima-compatible VAE loader") from e

When it happens

Trigger: Connecting a VAE field whose loaded model is e.g. AutoencoderKL (SD/SDXL VAE) or another unsupported class to the Anima ImageToLatents node's vae input.

Common situations: Selecting a mismatched VAE in a workflow (SDXL VAE with an Anima pipeline); a model-manager lookup returning the wrong model due to duplicate keys; workflows copied across pipelines with different model families.

Related errors

• Expected AutoencoderKL or FluxAutoEncoder, got {type(vae).__
• Expected ModelPatchRaw for LoRA '{lora.lora.key}', got {type
• Expected AutoencoderKLWan or FluxAutoEncoder, got {type(vae)
• Expected AutoencoderKLWan or FluxAutoEncoder for Anima VAE,
• Expected AutoencoderKLWan or FluxAutoEncoder, got {type(vae)

AI-assisted analysis of invoke-ai/InvokeAI@0b6a024f2f (2026-08-29). Data as JSON: /api/errors/7641f971452e876f. Report an issue: GitHub.