BerriAI/litellm · error · ValueError

Nova Canvas image edit requires an image input

Error message

Nova Canvas image edit requires an image input

What it means

Every Nova Canvas image edit fundamentally needs a source image. The helper _file_types_to_b64 encodes the OpenAI image input to base64 and raises ValueError when image is None, i.e. the request reached transformation without any image payload.

Source

Thrown at litellm/llms/bedrock/image_edit/amazon_nova_canvas_image_edit_transformation.py:134

        return {"taskType": "INPAINTING", "inPaintingParams": in_params}
    var_params: Final[dict[str, Any]] = {
        "images": [image_b64],
        "text": text,
    }
    if negative_text is not None:
        var_params["negativeText"] = negative_text
    if similarity_strength is not None:
        var_params["similarityStrength"] = similarity_strength
    return {
        "taskType": "IMAGE_VARIATION",
        "imageVariationParams": var_params,
    }


def _file_types_to_b64(image: FileTypes | None) -> str:
    """Encode OpenAI image input to base64 string for Nova Canvas."""
    if image is None:
        raise ValueError("Nova Canvas image edit requires an image input")
    if hasattr(image, "read") and callable(getattr(image, "read", None)):
        if hasattr(image, "seek"):
            image.seek(0)
        image_bytes: Final = image.read()
        return base64.b64encode(image_bytes).decode("utf-8")
    if isinstance(image, bytes):
        return base64.b64encode(image).decode("utf-8")
    if isinstance(image, str):
        return image
    if isinstance(image, os.PathLike):
        with open(image, "rb") as f:
            return base64.b64encode(f.read()).decode("utf-8")
    if isinstance(image, tuple):
        raise ValueError(
            "Nova Canvas image edit does not support tuple FileTypes. "
            "Pass a file-like object, bytes, or a base64-encoded string."
        )
    return base64.b64encode(bytes(image)).decode("utf-8")

View on GitHub (pinned to 6c2dcb801b)

Solutions

1. Include the source image as the 'image' file part of the multipart images/edits request.
2. Acceptable forms: file-like object, raw bytes, base64 string, or os.PathLike path (tuples are rejected separately).
3. If you meant generation from text only, call /v1/images/generations instead.

Example fix

# before
litellm.image_edit(model="bedrock/nova-canvas", prompt="a cat")  # no image
# after
with open("cat.png", "rb") as f:
    litellm.image_edit(model="bedrock/nova-canvas", prompt="a cat", image=f)

def has_image(image) -> bool:
    return image is not None and image != []

def is_usable_nova_image(v: object) -> bool:
    import os
    return (
        v is not None
        and (hasattr(v, "read") or isinstance(v, (bytes, str, os.PathLike)))
    )

When it happens

Trigger: POST /v1/images/edits on a nova-canvas model with no 'image' file part (only prompt), or litellm.image_edit(image=None) / empty list; also when a gateway drops the multipart image field.

Common situations: Client code written for DALL-E-style prompt-only edits; sending image as a JSON field instead of multipart file; test harness that omits fixtures.

Related errors

• OUTPAINTING requires either a mask image or a mask prompt. P
• Unsupported Amazon Nova Canvas taskType: {task_type!r}. Use
• Amazon Nova Canvas INPAINTING requires either maskPrompt or
• Nova Canvas image edit does not support tuple FileTypes. Pas
• Amazon Nova Canvas {task_type} requires a text prompt. Pass

AI-assisted analysis of BerriAI/litellm@6c2dcb801b (2026-08-15). Data as JSON: /api/errors/f49555bb974ed1f0. Report an issue: GitHub.