How do I use glm-vision-mcp-server?

1. Click on "Deploy Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@glm-vision-mcp-server Convert this design to HTML: https://example.com/design.png" That's it! The server will respond to your query, and you can continue using it as needed. Here is a step-by-step guide with screenshots.

glm-vision-mcp-server

by kira4094

Overview Schema Source Related Servers Discussions Admin

JavaScript

Hybrid

GLM-Vision MCP Server

npm version GitHub

Zhipu AI GLM 视觉模型 MCP Server，支持 GLM-5V-Turbo / GLM-4.6V 等，用于 Claude Code。

特性

🖼️ 支持本地图片文件路径和远程 URL
💬 自定义 prompt，适合 UI 截图→代码、视觉调试等场景
🧠 可选 Thinking 模式（复杂推理）
📏 200K 上下文 / 128K 最大输出
🏆 Design2Code 基准 94.8 分
⚡ 一行 npx 部署，无需手动安装

Related MCP server: Vision MCP

快速开始（npx）

Claude Code settings.json 中添加：

{
  "mcpServers": {
    "glm-5v": {
      "command": "npx",
      "args": ["-y", "glm-vision-mcp-server"],
      "env": {
        "ZAI_API_KEY": "your-zai-api-key"
      }
    }
  }
}

环境变量

变量	必填	默认值	说明
`ZAI_API_KEY`	✅	—	Z.AI API Key，去 z.ai 注册获取
`GLM_MODEL`	否	`glm-5v-turbo`	模型名，可换成 `glm-4.6v`、`glm-4.6v-flash`（免费）等

本地开发

git clone https://github.com/kira4094/glm-vision-mcp-server.git
cd glm-vision-mcp-server
npm install

本地路径注册到 Claude Code：

{
  "mcpServers": {
    "glm-5v": {
      "command": "node",
      "args": ["E:\\Projects\\Claude\\MCP\\GLM\\glm-vision-mcp-server\\src\\index.js"],
      "env": {
        "ZAI_API_KEY": "your-key-here"
      }
    }
  }
}

工具说明

`glm_5v_understand`

分析图片的核心工具。

参数	必填	说明
`image`	✅	本地路径 (`C:/img.png`) 或 URL (`https://...`)
`prompt`	✅	对图的指令。越具体越好
`detail`	否	精度: `auto` / `low` / `high`
`max_tokens`	否	最大输出（默认 4096，最大 128K）
`temperature`	否	采样温度（默认 1）
`thinking`	否	启用思考模式

使用示例

在 Claude Code 中直接问：

分析这张 UI 截图：C:\screenshot.png，描述它的布局和配色方案

把这张设计稿生成 HTML + Tailwind CSS：https://example.com/mockup.png

与 Doubao Vision 的对比

	GLM-5V-Turbo	Doubao Vision 1.6
API 提供商	Z.AI (智谱)	字节跳动
上下文	200K	各有千秋
Design2Code	94.8	—
价格	$1.20/M in · $4.00/M out	见豆包定价
特长	UI→代码、视觉编码	通用视觉理解

Available Tools

1 tool

glm_5v_understandA

Analyze an image using GLM-5V-Turbo (Zhipu AI's multimodal vision model). Supports local image files and remote URLs. Excels at: UI screenshot→code, design mockup analysis, visual debugging, chart/document understanding.

ParametersJSON Schema

Name	Required	Description	Default
`image`	Yes	Image source: local file path (e.g. C:/path/to/screenshot.png) or URL (https://...)
`detail`	No	Image detail level. 'high' for fine-grained UI analysis	auto
`prompt`	Yes	What to ask about the image. Be specific for best results. E.g.: 'Recreate this UI as HTML with Tailwind CSS'
`thinking`	No	Enable thinking mode for complex reasoning tasks
`max_tokens`	No	Maximum output tokens (max 128K)
`temperature`	No	Sampling temperature (0-2)

TDQS

A4/5.0

Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full burden of behavioral disclosure. It usefully states supported input types ('local image files and remote URLs'), which is beyond what the schema states. However, it does not mention output format, limitations, or side effects. This is adequate but not rich.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is three sentences, front-loaded with the core purpose, and includes a succinct list of use cases. Every sentence earns its place with no fluff or repetition.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a vision-analysis tool with no output schema, the description covers purpose, supported inputs, and primary use cases. It doesn't describe return format, but the 'analyze' verb implies textual output. Given no annotations, the description is fairly complete, though it could mention what the tool returns.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema provides 100% parameter coverage with detailed descriptions, setting a baseline of 3. The description adds context for the image parameter (local/remote) and prompt examples, but does not significantly exceed the schema's existing documentation. It adds marginal value.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly and specifically states the tool's function: 'Analyze an image using GLM-5V-Turbo'. It also lists concrete use cases (UI screenshot→code, design mockup analysis, visual debugging, chart/document understanding), making the purpose unambiguous even without sibling comparisons.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The 'Excels at' section provides clear guidance on when to use the tool—for UI screenshots, design mockups, visual debugging, and chart/document understanding. While it doesn't explicitly mention alternatives or exclusions, the use-case focus effectively implies suitable scenarios. No siblings exist to compare against.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

1 tool updatev1.0.0Aug 1, 2026
- First observedglm_5v_understand

Deploy Server

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

TDQS

A4.1/5.0

Scored 2026-08-01 06:50 across 1 tool

Disambiguation5/5

Only one tool exists, so there is no possibility of confusing it with other tools. The tool's purpose is clearly defined and unambiguous.

Naming Consistency5/5

The single tool name follows a clear pattern: model identifier (glm_5v) plus action verb (understand). Since there are no other tools, naming consistency is trivially perfect.

Tool Count3/5

With just one tool, the server feels minimal. While it serves a focused purpose, the count is on the low end of reasonable and could benefit from additional related capabilities.

Completeness5/5

For a vision-understanding server, the provided tool covers a broad range of tasks including UI analysis, chart understanding, and document processing. No obvious gaps exist for the stated domain.

Maintenance

ActivityInactive

ResponsivenessNo issues

Resources

Need Help?

Related Servers

Related MCP Connectors

lightgenOAuth
app.lightgen
Generate and edit images and create short videos inside Claude. Prepaid credits, no subscription.
Image & Video Processing Multimedia Processing
2
Writingmate MCPOAuth
ai.writingmate
Use AI models for chat, image, and video generation from Claude Code and other MCP hosts.
AI & Machine Learning Image & Video Processing
1
holiday-photo-plus A
io.github.holiday-forever
Show holiday photos from Vercel Blob inline in Claude chats.
Cloud Storage Entertainment & Media
12
Codex Imagem A
io.github.mcp-dir
Generate images with your own ChatGPT subscription (Plus, Pro or Team), without spending API credits
AI & Machine Learning Image & Video Processing
81

Related MCP Servers

GLM Vision Server
danilofalcao
A
license
B
quality
D
maintenance
Enables image analysis using GLM-4.5V's vision capabilities from Z.AI. Supports analyzing both local image files and URLs with customizable prompts and parameters.
Updated a year ago (2025-09-17 18:01 UTC)
1
7
MIT
Vision MCP
Image & Video Processing Multimedia Processing
i-richardwang
A
license
A
quality
D
maintenance
Enables image analysis and understanding using Vision Language Models through OpenAI-compatible APIs. Supports analyzing images from URLs or local files with custom prompts.
Updated 9 months ago (2025-12-26 05:40 UTC)
1
2
MIT
MCP Vision Server
Image & Video Processing AI & Machine Learning
coffe-d
A
license
A
quality
C
maintenance
Enables Claude Code to describe images and extract text using Kimi/Moonshot vision API. Supports local image files with customizable prompts.
Updated 4 months ago (2026-05-15 09:07 UTC)
2
MIT
glm-mcp-server
AI & Machine Learning Image & Video Processing
ifolin
A
license
Not graded
quality
D
maintenance
Enables image analysis in Claude Code using Zhipu AI's GLM-4.6V vision model.
Updated 6 months ago (2026-03-28 04:47 UTC)
5
MIT

GLM-Vision MCP Server

特性

快速开始（npx）

环境变量

本地开发

工具说明

glm_5v_understand

使用示例

与 Doubao Vision 的对比

Available Tools

TDQS

Tool Schema Changelog

TDQS

Maintenance

Resources

Related MCP Connectors

Related MCP Servers

GLM Vision Server

Vision MCP

MCP Vision Server

glm-mcp-server

`glm_5v_understand`