[Bug]使用deepseekv4 flash报400错误

[ltk415]

What happened / 发生了什么

使用deepseekv4 flash报400错误

Reproduce / 如何复现？

昨日正常使用，今早出现问题，我没有做任何改动。

AstrBot version, deployment method (e.g., Windows Docker Desktop deployment), provider used, and messaging platform used. / AstrBot 版本、部署方式（如 Windows Docker Desktop 部署）、使用的提供商、使用的消息平台适配器

mac系统，astrbot4.25.2，docker部署napcat

OS

macOS

Logs / 报错日志

[2026-06-05 07:43:05.870] [Core]
[WARN]
[v4.25.2] [runners.tool_loop_agent_runner:557]: Chat Model openai/deepseek-v4-flash request error: Error code: 400 - {'error': {'message': "Error from provider (DeepSeek): Messages with role 'tool' must be a response to a preceding message with 'tool_calls'", 'type': 'invalid_request_error', 'param': None, 'code': 'invalid_request_error'}}
Traceback (most recent call last):
File "/Applications/AstrBot.app/Contents/Resources/backend/app/astrbot/core/agent/runners/tool_loop_agent_runner.py", line 512, in _iter_llm_responses_with_fallback
async for attempt in retrying:
File "/Applications/AstrBot.app/Contents/Resources/backend/python/lib/python3.12/site-packages/tenacity/asyncio/init.py", line 170, in anext
do = await self.iter(retry_state=self._retry_state)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Applications/AstrBot.app/Contents/Resources/backend/python/lib/python3.12/site-packages/tenacity/asyncio/init.py", line 157, in iter
result = await action(retry_state)
^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Applications/AstrBot.app/Contents/Resources/backend/python/lib/python3.12/site-packages/tenacity/_utils.py", line 111, in inner
return call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/Applications/AstrBot.app/Contents/Resources/backend/python/lib/python3.12/site-packages/tenacity/init.py", line 393, in
self._add_action_func(lambda rs: rs.outcome.result())
^^^^^^^^^^^^^^^^^^^
File "/Applications/AstrBot.app/Contents/Resources/backend/python/lib/python3.12/concurrent/futures/_base.py", line 449, in result
return self.__get_result()
^^^^^^^^^^^^^^^^^^^
File "/Applications/AstrBot.app/Contents/Resources/backend/python/lib/python3.12/concurrent/futures/_base.py", line 401, in __get_result
raise self._exception
File "/Applications/AstrBot.app/Contents/Resources/backend/app/astrbot/core/agent/runners/tool_loop_agent_runner.py", line 516, in _iter_llm_responses_with_fallback
async for resp in self._iter_llm_responses(
File "/Applications/AstrBot.app/Contents/Resources/backend/app/astrbot/core/agent/runners/tool_loop_agent_runner.py", line 479, in _iter_llm_responses
yield await self.provider.text_chat(**payload)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Applications/AstrBot.app/Contents/Resources/backend/app/astrbot/core/provider/sources/openai_source.py", line 1286, in text_chat
) = await self._handle_api_error(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Applications/AstrBot.app/Contents/Resources/backend/app/astrbot/core/provider/sources/openai_source.py", line 1232, in _handle_api_error
raise e
File "/Applications/AstrBot.app/Contents/Resources/backend/app/astrbot/core/provider/sources/openai_source.py", line 1274, in text_chat
llm_response = await self._query(payloads, func_tool)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Applications/AstrBot.app/Contents/Resources/backend/app/astrbot/core/provider/sources/openai_source.py", line 639, in _query
completion = await self.client.chat.completions.create(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Applications/AstrBot.app/Contents/Resources/backend/python/lib/python3.12/site-packages/openai/resources/chat/completions/completions.py", line 2739, in create
return await self._post(
^^^^^^^^^^^^^^^^^
File "/Applications/AstrBot.app/Contents/Resources/backend/python/lib/python3.12/site-packages/openai/_base_client.py", line 1931, in post
return await self.request(cast_to, opts, stream=stream, stream_cls=stream_cls)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Applications/AstrBot.app/Contents/Resources/backend/python/lib/python3.12/site-packages/openai/_base_client.py", line 1716, in request
raise self._make_status_error_from_response(err.response) from None
openai.BadRequestError: Error code: 400 - {'error': {'message': "Error from provider (DeepSeek): Messages with role 'tool' must be a response to a preceding message with 'tool_calls'", 'type': 'invalid_request_error', 'param': None, 'code': 'invalid_request_error'}}
[2026-06-05 07:43:05.898] [Core]
[WARN]
[v4.25.2] [runners.tool_loop_agent_runner:494]: Switched from openai/deepseek-v4-flash to fallback chat provider: openai/mimo-v2.5

Are you willing to submit a PR? / 你愿意提交 PR 吗？

• Yes!

Code of Conduct

• I have read and agree to abide by the project's Code of Conduct。

[dosubot]

这个问题与 #8482 相同，是 DeepSeek V4 reasoning 模型在 tool loop 场景下的兼容性问题。

根本原因： DeepSeek V4 模型有时只返回 reasoning_content，而 content 和 tool_calls 均为空。当这条 assistant 消息被保存到对话历史后，后续请求会因为 tool message 找不到对应的 tool_calls 而返回 400 错误 [1]。

当前状态： PR #8483 已经提出修复方案（为这类消息设置 content="" 占位符），已在真实环境验证有效，但尚未合并到主分支 [2]。

临时解决方案：

• 执行 /new 清除当前对话历史，开启新对话可暂时规避
• 如果频繁复现，可考虑暂时关闭工具调用功能，或切换到非 DeepSeek V4 的模型

To reply, just mention @dosu.

---

Docs are dead. Just use Dosu.

[ltk415]

按照#8483 的方案修了之后，插件会报'NoneType' object has no attribute 'input_tokens'错误