BerriAI/litellm · error · Exception

Failed to upload file: {response.text}

Error message

Failed to upload file: {response.text}

What it means

The final PUT to the resumable upload URL returned a non-200/201 status, so the file content was not (fully) accepted by Gemini. The response body '{response.text}' gives the reason; a fresh upload session is required on retry since resumable URLs are single-use.

Source

Thrown at litellm/rag/ingestion/gemini_ingestion.py:319

            "X-Goog-Upload-Command": "upload, finalize",
        }

        verbose_logger.debug("Uploading file content (%s bytes)", len(file_content))

        client: Final = get_async_httpx_client(
            llm_provider=httpxSpecialProvider.RAG,
            params={"timeout": 300.0},  # Longer timeout for large files
        )
        response: Final = await client.put(
            upload_url,
            content=file_content,
            headers=headers,
        )

        if response.status_code not in [200, 201]:
            error_msg: Final = f"Failed to upload file: {response.text}"
            verbose_logger.error(error_msg)
            raise Exception(error_msg)

        # Parse response to get file/document ID
        try:
            response_data: Final = response.json()
            # The response should contain the document name or file reference
            file_id: Final = response_data.get("name", "") or response_data.get("file", {}).get("name", "")
            verbose_logger.debug("Upload complete. File ID: %s", file_id)
            return file_id
        except Exception as e:
            verbose_logger.warning("Could not parse upload response: %s", e)
            # Return a placeholder if we can't get the ID
            return "uploaded"

View on GitHub (pinned to 77b7c6c40c)

Solutions

1. Check the response text for the upload error and retry with a valid file.

When it happens

Trigger: Thrown at litellm/rag/ingestion/gemini_ingestion.py:319 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/9d0dfe2f5073d652. Report an issue: GitHub.