chroma-core/chroma · error · ValueError

Input documents cannot be empty

Error message

Input documents cannot be empty

What it means

GoogleGeminiEmbeddingFunction.__call__ rejects empty input before contacting the API. The Gemini embed_content endpoint requires at least one content item, and embedding zero documents is meaningless, so a falsy input (empty list/tuple) raises immediately.

Source

Thrown at chromadb/utils/embedding_functions/google_embedding_function.py:90

            project=project,
            location=location,
            http_options=types.HttpOptions(
                headers={"x-goog-api-client": f"chroma/{__version__}"}
            ),
        )

    def __call__(self, input: Documents) -> Embeddings:
        """
        Generate embeddings for the given documents.

        Args:
            input: Documents to generate embeddings for.

        Returns:
            Embeddings for the documents.
        """
        if not input:
            raise ValueError("Input documents cannot be empty")
        if not isinstance(input, (list, tuple)):
            raise ValueError("Input must be a list or tuple of documents")
        if not all(isinstance(doc, str) for doc in input):
            raise ValueError("All input documents must be strings")

        from google.genai.types import EmbedContentConfig

        config = EmbedContentConfig(
            task_type=self.task_type,
            output_dimensionality=self.dimension,
        )

        try:
            response = self.client.models.embed_content(
                model=self.model_name,
                contents=input,
                config=config,
            )

View on GitHub (pinned to aecdd12c8a)

Solutions

1. Short-circuit before embedding: if the batch is empty, skip the call or return []
2. Fix chunking logic so it never yields an empty final chunk (e.g. range(0, n, size))
3. Validate user-supplied query text is non-empty before calling collection.query

Example fix

# before
ef = GoogleGeminiEmbeddingFunction()
vecs = ef(batch)  # crashes when batch == []

# after
ef = GoogleGeminiEmbeddingFunction()
vecs = ef(batch) if batch else []

def safe_embed(ef, docs):
    if not docs:
        return []  # nothing to embed; skip the API entirely
    return ef(docs)

from typing import Any

def is_nonempty_batch(docs: Any) -> bool:
    return isinstance(docs, (list, tuple)) and len(docs) > 0

When it happens

Trigger: Calling ef([]) directly; passing an empty batch produced by chunking code whose last chunk is empty; query flows where the query_texts filter or user input produced zero items; collection.query with an empty documents/query_texts list reaching the embedding function.

Common situations: Batch ingestion loops that slice documents into fixed-size chunks and send a final empty remainder; upstream filters that legitimately return zero matches but are still passed to the embedding function; empty user search box submitted without validation.

Related errors

• Input must be a list or tuple of documents
• All input documents must be strings
• Invalid URL. Unrecognized protocol - {parsed.scheme}.
• Invalid URL. Seems that you are trying to pass URL as a host
• Record does not contain any non-None fields that can be embe

AI-assisted analysis of chroma-core/chroma@aecdd12c8a (2026-08-16). Data as JSON: /api/errors/21d64dc70753a788. Report an issue: GitHub.