multimodal-bridge
multimodal-bridge 是一个多模态能力桥：把 Qwen 的视觉理解（Qwen-VL）与图像生成（Qwen-Image）带给没有原生多模态能力的纯文本模型（如 DeepSeek）。它有两种形态、同一套后端： MCP Server（qwen_vision / qwen_generate 工具）：任何支持 MCP 的宿主（Claude Code、Kimi Code 等）直接挂载； DSH 插件（npm 包 dsh-multimodal-bridge，DeepSeek Harness bundle）：dsh plugin add 一行安装，含模型自动 fallback、尺寸自适应与图片结果卡片。
PluginVisionDeveloperNetworkAI ModelsVision / OCR
Verification
L2 · Structured
Security
Pending
Health
Active
Trust
Bronze
What it doesAI
Bridges Qwen-VL vision understanding and Qwen-Image generation to text-only models like DeepSeek.
- Adds Qwen-VL visual understanding to text-only LLMs
- Adds Qwen-Image generation for multimodal output
- Installable as MCP server or DSH plugin
AI-generated from the repo README — for reference only.
Installation
dsh plugin --profile web add multimodal-bridgeInstall method: npm · not yet tested in container (L3+)
Compatibility
| DSH Version | Status |
|---|---|
| not stated | Declared — not tested |
Requirements
- • Node.js: not stated
- • DSH: declared "not stated"
- • External credentials: none detected
Security Report
Automated scan, not manual review.
• Dependency vulnerabilities: — (pending)
• Suspicious permissions: — (pending)
• Hardcoded secrets: — (pending)
• Supply chain risks: — (pending)
⚠ Security audit scheduled — results will appear here after the next scan.
Activity
Last commit 2026-08-13 · activity: active
6 mo