BerriAI/litellm · error · ValueError

prefetched_source_data is required for Vertex AI video edit.

Error message

prefetched_source_data is required for Vertex AI video edit. Ensure get_video_edit_prefetch_params is called by the handler.

What it means

Raised by the Vertex edit request transform when prefetched_source_data is None. In normal operation the LiteLLM handler (litellm/llms/custom_httpx/llm_http_handler.py:7680) calls get_video_edit_prefetch_params, performs the fetchPredictOperation call, and passes the operation JSON back in. This ValueError therefore means the transform was invoked directly or via a code path that skipped the prefetch step.

Source

Thrown at litellm/llms/vertex_ai/videos/transformation.py:690

    def transform_video_edit_request(
        self,
        prompt: str,
        video_id: str,
        api_base: str,
        litellm_params: GenericLiteLLMParams,
        headers: dict,
        extra_body: dict[str, object] | None = None,
        prefetched_source_data: dict[str, Any] | None = None,
    ) -> tuple[str, dict]:
        """
        Build a predictLongRunning edit request from the pre-fetched source video.

        The actual fetchPredictOperation HTTP call is hoisted into the handler so
        it can use the shared async/sync httpx client instead of blocking the loop.
        """
        if prefetched_source_data is None:
            raise ValueError(
                "prefetched_source_data is required for Vertex AI video edit. "
                "Ensure get_video_edit_prefetch_params is called by the handler."
            )

        if not prefetched_source_data.get("done", False):
            raise ValueError("Source video generation is not complete yet. Check the video status before editing.")

        source_response: Final[_VeoOperationResponse] = prefetched_source_data.get("response", {})
        videos: Final = source_response.get("videos", [])
        if not videos:
            raise ValueError("No videos found in the completed operation. Cannot edit.")

        source_video: Final = videos[0]
        video_input: Final[dict[str, str]] = {}
        if "gcsUri" in source_video:
            video_input["gcsUri"] = source_video["gcsUri"]
        elif "bytesBase64Encoded" in source_video:
            video_input["bytesBase64Encoded"] = source_video["bytesBase64Encoded"]

View on GitHub (pinned to 77b7c6c40c)

Solutions

Use the standard litellm.video_edit(...) entry point so the handler performs the prefetch for you
If calling the transform directly, first call get_video_edit_prefetch_params, HTTP GET that URL, and pass response.json() as prefetched_source_data

Example fix

# before
url, body = config.transform_video_edit_request(video_id, prompt, api_base, litellm_params, headers)
# ValueError: prefetched_source_data is required

# after
p_url, p_body = config.get_video_edit_prefetch_params(video_id, api_base, litellm_params, headers)
resp = httpx.get(p_url, headers=headers)  # fetchPredictOperation
url, body = config.transform_video_edit_request(
    video_id, prompt, api_base, litellm_params, headers,
    prefetched_source_data=resp.json(),
)

Defensive patterns

Strategy: validation

Validate before calling

def build_edit_request(config, video_id, prompt, api_base, litellm_params, headers, http_get):
    p_url, _ = config.get_video_edit_prefetch_params(video_id, api_base, litellm_params, headers)
    prefetched = http_get(p_url, headers).json()  # fetchPredictOperation
    return config.transform_video_edit_request(
        video_id, prompt, api_base, litellm_params, headers,
        prefetched_source_data=prefetched,
    )

Type guard

from typing import Any

def has_prefetched_source_data(kwargs: dict[str, Any]) -> bool:
    return isinstance(kwargs.get("prefetched_source_data"), dict)

Try / catch

try:
    url, body = config.transform_video_edit_request(video_id, prompt, api_base, lp, headers, prefetched_source_data=prefetched)
except ValueError as e:
    if "prefetched_source_data is required" in str(e):
        raise RuntimeError("handler bug: prefetch step was skipped") from e
    raise

Prevention

Prefer the public litellm.video_edit entry point over calling transformations directly
If invoking transforms manually, always pair get_video_edit_prefetch_params + HTTP fetch + pass the JSON back in
Assert prefetched_source_data is not None before calling the transform in custom handlers

When it happens

Trigger: Calling config.transform_video_edit_request(...) manually in custom code; a forked/modified handler that builds the edit request without first fetching the source operation; provider interface changes where a handler forgets the prefetch contract.

Common situations: Writing a custom video handler that mimics litellm's; upgrading litellm versions where the handler/prefetch contract shifted; unit tests invoking the transformation in isolation.

Understand the failure class

Background: "missing required argument" and "the following required arguments were not provided": what required-argument errors mean and how to fix them — this error's family across 20 libraries.

Related errors

AI-assisted analysis of BerriAI/litellm@77b7c6c40c (2026-08-18). Data as JSON: /api/errors/7fc38543c32f4cbf. Report an issue: GitHub.