Skip to content

[Feature Request] MiniMax M2.7 模型支持图片输入 (Vision Input Support) #18

@WwNeXst

Description

@WwNeXst

问题描述

MiniMax M2.7 / M2.7 Highspeed 模型不支持图片输入(`input: ["text"]`),导致在使用 OpenClaw 等 MCP 客户端时,用户无法直接发送截图——图片在网关层就被丢弃。

现状

  • MiniMax M2.7 系列:`input: ["text"]`(不支持图片)
  • 竞品 Kimi K2.5:`input: ["text", "image"]`(支持图片)
  • MiniMax 提供了 `minimax-coding-plan-mcp` 中的 `understand_image` 工具支持图片理解,但该工具无法绕过 MCP 客户端的网关图片过滤层

期望

MiniMax M2.7 及以上模型增加图片输入支持(`input: ["text", "image"]`),与 moonshot/kimi-k2.5 等竞品看齐,提升在 Agent / MCP 场景下的用户体验。

相关链接

环境

  • Model: minimax/m2.7-highspeed
  • MCP: minimax-coding-plan-mcp (含 understand_image 工具)

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions