### Check for existing issues

- [x] I have searched the existing issues and checked that my issue is not a duplicate.

### What happened?

An unexpected blank assistant message emerged in the Litellm AI Gateway's connection with Deepseek's server's chat completion API while I was using codex, which used the `/responses` endpoint of the Litellm AI Gateway and the model `deepseek v4 flash`. This raised an `HTTP 400` error from Deepseek's server: `{"error":{"message":"An assistant message with 'tool_calls' must be followed by tool messages responding to each 'tool_call_id'. (insufficient tool messages following tool_calls message)","type":"invalid_request_error","param":null,"code":"invalid_request_error"}}. Received Model Group=deepseek-v4-flash Available Model Group Fallbacks=None` Involved messages were: `[{"tool_calls": [{"id": "call_00_spr6ZlhqFv74vJEmiDbJ8400","type": "function","function": {"name": "shell_command","arguments": "{\"command\": \"Get-ChildItem -Path \\\"D:\\\\626工程\\\" -Depth 0\", \"timeout_ms\":5000}"},"index": 0}],"role": "assistant"},{"role": "assistant","content": [{"type": "text","text": ""}]}]`

### Steps to Reproduce

1. Configure the Litellm AI Gateway …