Skip to content

[Bug] DeepSeek V4模型无法调用工具 #7855

@liushuill

Description

@liushuill

What happened / 发生了什么

DeepSeek V4的两个模型,无法调用工具(如定时提醒能力)。日志中反馈为:”deepseek-v4-flash 不支持函数工具调用,已自动去除,不影响使用。“

Reproduce / 如何复现?

1.更新DeepSeek的模型列表,切换为新模型。
2.在aiocqhttp平台(如OneBot)进行对话(在旧的对话中),出现参数错误,报告错误为:
Error code: 400 - {'error':{'message': 'Thereasoning_content in the thinking mode must be passed back to the API . ' , 'type': 'invalid_request_error', 'param': None, 'code': 'invalid_request_error'}}
3.删除该对话,开启新对话。可以正常与LLM交流,但在调用工具时,日志显示deepseek-v4-flash 不支持函数工具调用,已自动去除,不影响使用。
4.切换deepseek-v4-pro ,上述情况依旧

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

AstrBot:版本:4.23.5-4.23.6
部署方式:Windows Docker Desktop
提供商:Ubuntu 24.04 +wsl
消息平台适配器:WebChat + aiocqhttp

OS

Linux

Logs / 报错日志

[22:25:13.050] [Core] [INFO] [core.event_bus:61]: [Webchat] [webchat(webchat)] LSLL/LSLL: 嗯,现在Astrbot更新了,你能试着你自己给我设置一个提醒吗?

就设置23:00提醒我睡觉的

[22:25:14.412] [Core] [INFO] [respond.stage:183]: Prepare to send - LSLL/LSLL:

[22:25:14.412] [Core] [INFO] [respond.stage:199]: 应用流式输出(webchat)

[22:25:21.012] [Core] [INFO] [sources.openai_source:1131]: deepseek-v4-flash 不支持函数工具调用,已自动去除,不影响使用。

[22:25:27.682] [Core] [WARN] [v4.23.6] [runners.tool_loop_agent_runner:821]: skills_like tool re-query returned no tool calls; fallback to assistant response.

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

  • Yes!

Code of Conduct

Metadata

Metadata

Assignees

No one assigned

    Labels

    area:providerThe bug / feature is about AI Provider, Models, LLM Agent, LLM Agent Runner.bugSomething isn't working

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions