BerriAI/litellm · error · ProxyException

context_window_size must be non-negative

Error message

context_window_size must be non-negative

What it means

Range guard on the auto-router classifier default-prompt endpoint: the context_window_size query parameter was negative. It feeds the classifier prompt's sizing, so only non-negative values are accepted.

Source

Thrown at litellm/proxy/management_endpoints/model_management_endpoints.py:2244

) -> AutoRouterClassifierDefaultPromptResponse:
    """
    Get the default classifier system prompt, so the dashboard's prompt editor can prefill it.

    The prompt's closing line depends on whether prior conversation turns are quoted to the
    classifier, its tier bullets are named by the router's tier_labels, and its calibration examples
    come from the router's classification rubric, so the caller passes all three to get the text that router
    would actually send rather than a rubric it does not use.

    Parameters:
    - context_window_size: int - The router's classifier_context_window_size. Defaults to the
      built-in default.
    - tier_labels: str | None - The router's tier_labels as a JSON object of canonical tier name to
      display name, e.g. `{"SIMPLE": "Cheap"}`. Omit or pass an empty object for the default names.
    - classification_rubric: ClassificationRubric | None - The router's
      classifier_llm_config.classification_rubric. Omit for the default.
    """
    if context_window_size < 0:
        raise ProxyException(
            message="context_window_size must be non-negative",
            type=ProxyErrorTypes.bad_request_error,
            code=status.HTTP_400_BAD_REQUEST,
            param="context_window_size",
        )

    labeled_tiers: Final = _labeled_tiers_from_query(tier_labels)
    return AutoRouterClassifierDefaultPromptResponse(
        system_prompt=(
            classification_system_prompt(context_window_size, classification_rubric=classification_rubric)
            if labeled_tiers is None
            else classification_system_prompt(
                context_window_size, labeled_tiers=labeled_tiers, classification_rubric=classification_rubric
            )
        )
    )

View on GitHub (pinned to 77b7c6c40c)

Solutions

1. Set context_window_size to a non-negative integer.

When it happens

Trigger: Thrown at litellm/proxy/management_endpoints/model_management_endpoints.py:2244 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of BerriAI/litellm@77b7c6c40c (2026-08-18). Data as JSON: /api/errors/df11238923e5c21c. Report an issue: GitHub.