Claude Code Integration with Seedream MCP

📘 Overview：Claude Code MCP Overview →

Generating images with Chinese text using English models often results in garbled characters—words like "新年快乐" produced by Midjourney are completely unrecognizable. Seedream, developed by ByteDance, specifically addresses this issue and is currently the most accurate AI drawing model for rendering Chinese text. After receiving Claude Code, you can generate images with Chinese text in just one command in the terminal.

Obtain API Token

Before using the Seedream MCP Server, you need to prepare an Ace Data Cloud API Token. The method for obtaining it is consistent with the Claude Code VS Code configuration tutorial:

  1. Open the Ace Data Cloud Console - Application List to get your API Token for backup.
  2. If you are not logged in or registered, you will be automatically redirected to the login page; after logging in or registering, you will be returned to the current page.
  3. The first application will come with a free quota, allowing you to experience the Claude Code MCP service for free.

One Token can be used for all MCP Servers provided by AceData Cloud, and there is no need to apply separately for Seedream. The documentation and screenshots suggest displaying only a masked version, such as 3b78cc40dd3b43db806a4300...., and do not paste the full Token into public repositories, issues, screenshots, or chat records.

Using the web version of claude.ai or Claude Desktop? They support OAuth one-click authorization, eliminating the need to manually enter the Token. For details, see the Claude.ai / Desktop tutorial for Seedream MCP.

Configure Claude Code

Choose one of the following three options. -H must be uppercase; the lowercase -h is --help; please replace the Token after Authorization with the actual value you copied from the console.

For Current Project Only: local

Suitable for trial use, this configuration is bound only to the project directory where you run the command. Claude Code will write the record into the local ~/.claude.json, including the current project path.

claude mcp add seedream --transport http https://seedream.mcp.acedata.cloud/mcp \
  -H "Authorization: Bearer 3b78cc40dd3b43db806a4300...." \
  -s local

Usable for All Projects: user

Suitable for frequent use of Seedream MCP across multiple projects. The configuration is written into the local ~/.claude.json user-level configuration, allowing you to see it in any project when opening Claude Code.

claude mcp add seedream --transport http https://seedream.mcp.acedata.cloud/mcp \
  -H "Authorization: Bearer 3b78cc40dd3b43db806a4300...." \
  -s user

Shared with Project: project

Suitable for team projects. The configuration is written into the .mcp.json in the current project root directory, which can be submitted to a private repository for teammates to reuse; do not submit the real Token to public repositories, and it is recommended to change it to an environment variable placeholder or have each person add it locally.

claude mcp add seedream --transport http https://seedream.mcp.acedata.cloud/mcp \
  -H "Authorization: Bearer 3b78cc40dd3b43db806a4300...." \
  -s project

When the project-level configuration is first read by Claude Code, it may display Pending approval, and you need to confirm trust in this project configuration within the Claude Code session. This is a normal security prompt.

Actual Running Results

Below is the masked output after actual execution in an isolated temporary directory using Claude Code 2.1.158. The command uses the real AceData Cloud Token, and the Token in the output has been replaced with 3b78cc40dd3b43db806a4300...., and the temporary configuration has been deleted.

Added HTTP MCP server seedream with URL: https://seedream.mcp.acedata.cloud/mcp to local config
seedream: https://seedream.mcp.acedata.cloud/mcp (HTTP) - ✓ Connected

The behavior of the three scopes has also been tested:

luma-local: https://luma.mcp.acedata.cloud/mcp (HTTP) - ✓ Connected
luma-user: https://luma.mcp.acedata.cloud/mcp (HTTP) - ✓ Connected
luma-project: https://luma.mcp.acedata.cloud/mcp (HTTP) - ⏸ Pending approval (run `claude` to approve)

To confirm the local configuration, you can run:

claude mcp list

Seeing seedream: https://seedream.mcp.acedata.cloud/mcp (HTTP) - ✓ Connected indicates that Seedream MCP has been successfully integrated.

Particularly Reliable for Chinese Scenarios

After configuration is complete, return to the Claude Code session and directly use natural language to call Seedream:

Images with Chinese Text

The text is clear, correct, and aesthetically pleasing:

Generate a promotional poster with the title "Developer Conference 2025" and the subtitle "AI Drives the Future," in a blue tech style.

Chinese Aesthetic

For an app targeting the domestic market, wanting a Chinese-style landing page:

Generate a new Chinese-style illustration with distant green mountains and clear waters, and nearby glowing digital particles flowing, blending technology and tradition.

Editing Text in Existing Images

Change "v1.0" on this poster to "v2.0," keeping everything else the same.

Tool List

Tool Description
seedream_generate_image Text to image generation
seedream_edit_image Edit existing images (change text, change elements)