add visual understanding capabilities-Alibaba Cloud Model Studio(Model Studio)-阿里云帮助中心

Prerequisites

You have subscribed to Coding Plan. See Getting started.
You have configured the connection in your Coding Plan tool and can start conversations. See Clients and developer tools.

Visual support status

Model	Visual support	Description
qwen3.6-plus qwen3.5-plus kimi-k2.5	Yes	No extra configuration required. Pass images directly.
qwen3-max-2026-01-23 qwen3-coder-next qwen3-coder-plus glm-5 glm-4.7 MiniMax-M2.5	No	Requires a Skill or Agent for vision.

Method 1: Use a visual model directly (recommended)

qwen3.6-plus, qwen3.5-plus, and kimi-k2.5 have built-in vision support and are recommended if you frequently work with images.

Tool	How to switch models
Claude Code	`/model qwen3.6-plus`, `/model qwen3.5-plus`, or `/model kimi-k2.5`
OpenCode	`/models` → Select `qwen3.6-plus`, `qwen3.5-plus`, or `kimi-k2.5`
Qwen Code	`/model` → Select `qwen3.6-plus`, `qwen3.5-plus`, or `kimi-k2.5`

To switch models in other tools, see Clients and developer tools. After switching, reference image paths or drag and paste images into the conversation.

Method 2: Add visual capabilities using a Skill or Agent

To process images using models without vision support, such as glm-5 and MiniMax-M2.5, configure a Skill or Agent.

Claude Code

Add a Skill

In your project directory, create a skills/image-analyzer directory in the .claude folder:

mkdir -p .claude/skills/image-analyzer

In this directory, create a SKILL.md file with the following content:

---
name: image-analyzer
description: Helps models without visual capabilities understand images. Use this skill to analyze image content, extract information, text, and UI elements from an image, or understand any visual content such as screenshots, charts, or architecture diagrams. Pass the image path to get a description.
model: qwen3.6-plus
---
qwen3.6-plus has visual understanding capabilities. Use the qwen3.6-plus model directly for image understanding.

Directory structure:

.claude/
└── skills/
    └── image-analyzer/
        └── SKILL.md

Get started
1. In your project directory, run claude to start Claude Code, and then run /model glm-5 to switch to glm-5.
2. Download aliyun.png to your project directory and ask the following question: Load the image-analyzer skill and describe the information in the aliyun.png banner. The reply is similar to the following:

OpenCode

Add an Agent

In your project directory, create an agents directory in the .opencode folder:

mkdir -p .opencode/agents

In this directory, create an image-analyzer.md file with the following content:

Note

The model field must use the provider and model name defined in the OpenCode config file. For example, based on the configuration in the OpenCode documentation, the value is bailian-coding-plan/qwen3.6-plus.

---
description: Analyzes images using a vision-capable model. Use this agent when the user needs to understand image content, extract information from screenshots, diagrams, UI mockups, or any visual content. Invoke with @image-analyzer followed by the image path and your question.
mode: subagent
model: bailian-coding-plan/qwen3.6-plus
tools:
  write: false
  edit: false
---
You have vision capabilities. Analyze the provided image and return a clear, structured description focused on what the user is asking about.

Directory structure:

.opencode/
└── agents/
    └── image-analyzer.md

Get started
1. In your project directory, run opencode to start OpenCode, and then switch to glm-5.
2. Download aliyun.png to your project directory. Use the at sign to invoke image-analyzer and ask the following question: @image-analyzer, describe the information in the aliyun.png banner. The reply is similar to the following:

FAQ

Why can't OpenCode + a vision model understand images?

Cause: OpenCode does not enable model visual capabilities by default. You must declare the modalities parameter in the config file.

Solution: In the model definition of the OpenCode config file, add the modalities field and set input to ["text", "image"]:

Replace sk-sp-xxx with your Coding Plan API key.

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "bailian-coding-plan-test": {
      "npm": "@ai-sdk/anthropic",
      "name": "Model Studio Coding Plan",
      "options": {
        "baseURL": "https://coding.dashscope.aliyuncs.com/apps/anthropic/v1",
        "apiKey": "sk-sp-xxx"
      },
      "models": {
        "qwen3.6-plus": {
          "name": "Qwen3.6 Plus",
          "modalities": {
            "input": [
              "text",
              "image"
            ],
            "output": [
              "text"
            ]
          },
          "options": {
            "thinking": {
              "type": "enabled",
              "budgetTokens": 1024
            }
          }
        },
        "qwen3.5-plus": {
          "name": "Qwen3.5 Plus",
          "modalities": {
            "input": [
              "text",
              "image"
            ],
            "output": [
              "text"
            ]
          },
          "options": {
            "thinking": {
              "type": "enabled",
              "budgetTokens": 1024
            }
          }
        },
        "kimi-k2.5": {
          "name": "Kimi K2.5",
          "modalities": {
            "input": [
              "text",
              "image"
            ],
            "output": [
              "text"
            ]
          },
          "options": {
            "thinking": {
              "type": "enabled",
              "budgetTokens": 1024
            }
          }
        }
      }
    }
  }
}

Why can't OpenClaw + a vision model understand images?

Cause: OpenClaw uses the input field in the config file to determine whether a model supports vision.

Solution:

In the ~/.openclaw/openclaw.json config file, ensure the model definition includes the "input": ["text", "image"] field.

{
  "models": {
    "mode": "merge",
    "providers": {
      "bailian": {
        "baseUrl": "https://coding.dashscope.aliyuncs.com/v1",
        "apiKey": "YOUR_API_KEY",
        "api": "openai-completions",
        "models": [
          {
            "id": "qwen3.6-plus",
            "name": "qwen3.6-plus",
            "reasoning": false,
            "input": ["text", "image"],
            "cost": { "input": 0, "output": 0, "cacheRead": 0, "cacheWrite": 0 },
            "contextWindow": 1000000,
            "maxTokens": 65536
          },
          {
            "id": "qwen3.5-plus",
            "name": "qwen3.5-plus",
            "reasoning": false,
            "input": ["text", "image"],
            "cost": { "input": 0, "output": 0, "cacheRead": 0, "cacheWrite": 0 },
            "contextWindow": 1000000,
            "maxTokens": 65536
          },
          {
            "id": "kimi-k2.5",
            "name": "kimi-k2.5",
            "reasoning": false,
            "input": ["text", "image"],
            "cost": { "input": 0, "output": 0, "cacheRead": 0, "cacheWrite": 0 },
            "contextWindow": 262144,
            "maxTokens": 32768
          }
        ]
      }
    }
  },
  "agents": {
    "defaults": {
      "model": {
        "primary": "bailian/qwen3.6-plus"
      },
      "models": {
        "bailian/qwen3.6-plus": {},
        "bailian/qwen3.5-plus": {},
        "bailian/kimi-k2.5": {}
      }
    }
  },
  "gateway": {
    "mode": "local"
  }
}

After modifying the configuration, clear the OpenClaw model cache and restart OpenClaw. Otherwise, the old configuration remains in effect.
```
rm ~/.openclaw/agents/main/agent/models.json
openclaw gateway restart
```