[BUG] Switching to mimo-v2.5-pro fails after reading images in mimo-v2.5
Status Open
Maintainer reply None cached
Activity 11 comments · opened May 26, 2026
Preflight Checklist
- [x] I have searched existing issues and this hasn't been reported yet
- [x] This is a single bug report (please file separate reports for different bugs)
- [x] I am using the latest version of Claude Code
What's Wrong?
在使用 Claude Code 插件时,如果在 mimo-v2.5 模型下读取了图片,之后切换到 mimo-v2.5-pro 模型,该模型将完全无法响应任何输入,即使只是简单的"你好"也会报错:切回 mimo-v2.5 后立即恢复正常。
复现步骤:
选择 mimo-v2.5 模型
使用 Read 工具读取一张图片
使用 /model 命令切换到 mimo-v2.5-pro
输入任何内容
对比测试:
同样的操作在 Cherry Studio(另一个 Claude Code 客户端)中可以正常切换,说明问题不在模型本身,而是 Claude Code 插件在处理包含图片的对话历史时存在问题。
What Should Happen?
切换到 mimo-v2.5-pro 后,应该能够正常对话,忽略或正确处理之前对话中的图片数据,而不是报错
Error Messages/Logs
There's an issue with the selected model (mimo-v2.5-pro[1m][1m]). It may not exist or you may not have access to it. Run --model to pick a different model.
Steps to Reproduce
1.新建一个 Claude Code 对话
2.选择 mimo-v2.5 模型
3.使用 Read 工具读取一张图片(例如:Read "path/to/image.png")
4.使用 /model 命令切换到 mimo-v2.5-pro
5.输入任何内容(如"你好")
6.观察报错
Claude Model
Other
Is this a regression?
I don't know
Last Working Version
_No response_
Claude Code Version
VSCode 插件版
Platform
Anthropic API
Operating System
Windows
Terminal/Shell
Windows Terminal
Additional Information
测试环境:
- 操作系统:Windows 11
- 编辑器:VSCode + Claude Code 插件
- 模型供应商:CC Switch (第三方工具)
对比测试:同样的操作在 Cherry Studio 中可以正常切换模型,
说明问题与 Claude Code 插件处理图片历史的方式有关。
11 Comments
模型一直是mimo v2.5 pro也会失败,只要碰到截图就失败。复现了2次
只要用v2.5pro模型识别了图片,就会失败。
大佬,有没有解决方案呀?
https://github.com/Mriestac/mimo-image-recognition-mcp
可以试试
https://platform.xiaomimimo.com/docs/zh-CN/quick-start/model
https://platform.xiaomimimo.com/docs/zh-CN/usage-guide/multimodal-understanding/image-understanding
因为mimo-v2.5-pro本身就不支持图片识别,官方文档写的图片理解当前仅支持 mimo-v2.5,mimo-v2-omni 模型。
<img width="576" height="366" alt="Image" src="https://github.com/user-attachments/assets/2e64f34c-f410-48c7-8273-b081b6313d6f" />
<img width="1896" height="868" alt="Image" src="https://github.com/user-attachments/assets/dd382a92-2760-4ab9-975e-b1f17df97401" />
+1, I'm also affected by this issue.
Environment:
Reproduction: Same as described — reading an image while on mimo-v2.5, then switching to mimo-v2.5-pro via triggers the error:
This is a significant usability issue for users relying on third-party models with vision capabilities. The model switch should either:
Hoping this can be prioritized. Thanks!
quick fix is to add on your session prompt that avoiding image reading
@WardLu deepseek v4 pro 也不支持识图,GLM 5.2 也不支持识图,但是他们不会报错,会告诉用户自己无法识图,或者尝试调用识图MCP。不是支持不支持的问题,是返回什么参数的问题
There's an issue with the selected model (mimo-v2.5-pro[1m]). It may not exist or you may not have
access to it. Run /model to pick a different model. 同样的问题,用mimo-v2.5就没有问题,希望修复
这是来自QQ邮箱的假期自动回复邮件。你好,我最近正在休假中,无法亲自回复你的邮件。我将在假期结束后,尽快给你回复。
+1 i have the same problem