BerriAI/litellm · error · ValueError

Unsupported image type for Vertex AI Imagen image edit.

Error message

Unsupported image type for Vertex AI Imagen image edit.

What it means

Inside _read_all_bytes, when the value is a list or tuple the code recurses into the first non-None element; if every element is None (or the container is empty) there is nothing to read and this ValueError is raised. It means a container reached the byte reader but carried no usable payload — distinct from the top-level 'image is None' check in transform_image_edit_request.

Source

Thrown at litellm/llms/vertex_ai/image_edit/vertex_imagen_transformation.py:298

                    "maskMode": "MASK_MODE_USER_PROVIDED",
                    "dilation": 0.03,  # Default dilation value (not configurable via OpenAI API)
                },
            }
            reference_images.append(mask_reference)

        return reference_images

    def _read_all_bytes(self, image: Any, depth: int = 0, max_depth: int = DEFAULT_MAX_RECURSE_DEPTH) -> bytes:
        if depth > max_depth:
            raise ValueError(
                f"Max recursion depth {max_depth} reached while reading image bytes for Vertex AI Imagen image edit."
            )

        if isinstance(image, (list, tuple)):
            for item in image:
                if item is not None:
                    return self._read_all_bytes(item, depth=depth + 1, max_depth=max_depth)
            raise ValueError("Unsupported image type for Vertex AI Imagen image edit.")

        if isinstance(image, dict):
            for key in ("data", "bytes", "content"):
                if key in image and image[key] is not None:
                    value = image[key]
                    if isinstance(value, str):
                        try:
                            return base64.b64decode(value)
                        except Exception:
                            continue
                    return self._read_all_bytes(value, depth=depth + 1, max_depth=max_depth)
            if "path" in image:
                return self._read_all_bytes(image["path"], depth=depth + 1, max_depth=max_depth)

        if isinstance(image, bytes):
            return image
        if isinstance(image, bytearray):
            return bytes(image)

View on GitHub (pinned to 77b7c6c40c)

Solutions

Filter Nones and empties out of image lists before calling litellm
Replace failed downloads with a hard error instead of None placeholders
Validate that at least one element is bytes/str-in-dict/file-like before sending

Example fix

# before
resp = litellm.image_edit(
    model='vertex_ai/imagen-3.0-capability-001',
    prompt='edit',
    image=[failed_download, None],  # both None-ish -> raises
)

# after
imgs = [img for img in downloads if img is not None]
if not imgs:
    raise ValueError('all image downloads failed')
resp = litellm.image_edit(
    model='vertex_ai/imagen-3.0-capability-001',
    prompt='edit',
    image=imgs,
)

Defensive patterns

Strategy: validation

Validate before calling

def has_payload(container) -> bool:
    return any(item is not None for item in (container or []))

if isinstance(image, (list, tuple)):
    assert has_payload(image), 'image container has no usable entries'

Try / catch

try:
    resp = litellm.image_edit(model='vertex_ai/imagen-...', prompt=p, image=image)
except ValueError as e:
    if 'Unsupported image type' in str(e) and isinstance(image, (list, tuple)):
        raise ValueError('image list contained only None entries') from e
    raise

Prevention

Filter Nones from every image list before it reaches litellm
Never use [None] placeholders to satisfy list-typed parameters
Log when image downloads fail so empty containers are visible upstream

When it happens

Trigger: image=[None] or image=[None, None] passed as the images list; the mask optional param set to an empty tuple (); a dict {'data': [None]} recursing into a None-only list.

Common situations: Placeholder lists filled with None by an earlier failed download; default arguments like image=[None] used to satisfy a type signature; multi-upload forms where every file failed to read and None was substituted.

Related errors

AI-assisted analysis of BerriAI/litellm@77b7c6c40c (2026-08-18). Data as JSON: /api/errors/c6494e340053fb17. Report an issue: GitHub.