Wan27 MCP — AI Image and Video Generation

io.github.NotUsingGithub由 GitHub 帳號 NotUsingGithub 發布v0.1.3更新於 Oct 24, 2026

Generate AI images and videos, estimate credits and track tasks. Paid Wan27 subscription required.

已驗證STDIO僅桌面媒體與設計AI 與機器學習

概覽

AI 產生的概覽

Wan27 MCP 讓助理在付費 Wan27 帳號上生成 AI 圖片與影片、查看模型選項、估算點數消耗並追蹤任務。

可以做什麼
  • 查詢帳號狀態與點數餘額
  • 列出並檢視可用的圖片與影片模型及其參數
  • 在正式生成前估算請求的點數花費
  • 生成圖片與影片,包括以上傳首幀進行圖生影片
  • 列出任務並取得單一任務的狀態或結果
適用情境
適合希望助理透過 Wan27 訂閱產出 AI 圖片或短影片,並在消耗點數前先了解價格的情境。也方便在同一帳號下查看任務進度並下載完成的素材。
執行需求
透過 npx 從 npm 套件 @wan27/cli 以 stdio 在本機執行,或以 Docker 容器執行。需要 macOS 或 Linux(含 WSL)上的 Node 22+。需要有效的付費 Wan27 訂閱,並用 CLI 的 auth login 指令登入一次;也可透過 WAN27_ACCESS_TOKEN 提供預發的權杖。憑證保存在 macOS 鑰匙圈或私有檔案中,目錄可用 WAN27_CONFIG_DIR 變更。
安裝前請注意
生成會依與網站相同的價格消耗 Wan27 網站帳戶中的實際點數,且存取需要付費訂閱;僅購買點數包、免費帳號與未付費試用都不符合資格。助理可以建立產生費用的任務,建議先查看帳號狀態與估算。憑證會保存在磁碟或鑰匙圈中,透過 WAN27_ACCESS_TOKEN 傳入或寫入容器映像的權杖需妥善保護;登出或變更網站密碼會撤銷連線。

安裝

在 SourceWeft 中

  1. 開啟 儀表板中的 Wan27 MCP — AI Image and Video Generation,將其新增到工作區。
  2. 為需要使用其工具的對話啟用該服務。

Desktop only,透過 STDIO。 STDIO 服務會啟動本機處理程序,因此需要 SourceWeft 桌面主機。

其他 MCP 客戶端

參照 儲存庫 中的啟動說明。

README

Wan27 MCP and CLI — AI Image and Video Generation

Generate images and videos from your terminal or AI assistant with your wan2-7.io account. MCP and CLI spend the same credit balance as the website, at the same prices. Credits are topped up on the website.

Access requires an active paid subscription. Credit packs alone, free accounts and unpaid trials do not qualify. A subscription with auto-renew turned off keeps access until its paid period ends. Upgrade at https://wan2-7.io/pricing/.

Node 22+, macOS or Linux (including WSL).

Install

sh
npm install -g @wan27/cli

Use

sh
wan27 auth loginwan27 accountwan27 modelswan27 estimate --input request.jsonwan27 create --input request.json --waitwan27 download <taskId> --output result.mp4wan27 auth logout

request.json examples:

json
{"modelId":"wan27-text-to-video","prompt":"A paper boat on a rainy street","durationSeconds":5,"resolution":"720p"}
json
{"modelId":"wan2.7-image","prompt":"A paper crane on a desk","resolution":"2K"}

For image-to-video, upload the first frame and pass its URL:

sh
wan27 upload frame.png        # {"url":"https://..."}
json
{"modelId":"wan27-image-to-video","prompt":"The crane takes off","imageUrl":"https://..."}

wan27 model <modelId> prints every option and its JSON Schema.

create prints a request ID on stderr. Re-running it with --request-id <id> after a network error returns the same task and is never charged twice. Failed tasks are refunded automatically. A wait timeout only stops waiting; the task keeps running.

Credentials are stored in the macOS Keychain, or with --credential-store file in a private file (directory 0700, file 0600). WAN27_CONFIG_DIR changes that directory. WAN27_ACCESS_TOKEN accepts a pre-issued API token for automation. Logging out revokes this connection; changing your website password revokes every connection.

Connect an assistant (local MCP)

Sign in once with wan27 auth login, then register the stdio server:

sh
codex mcp add wan27 -- wan27 mcpclaude mcp add wan27 -- wan27 mcp

Tools: account_status, list_models, get_model, estimate_generation, generate_image, generate_video, list_tasks, get_task.

After signing in, clients that accept JSON configuration can launch the same stdio server with npx:

json
{  "mcpServers": {    "wan27": {      "command": "npx",      "args": ["-y", "@wan27/cli", "mcp"]    }  }}

Read account status and estimate the cost before generating. Image and video generation spend your website credits.

Documentation and support

Run this public distribution

This repository contains the readable JavaScript executable for @wan27/[email protected]. The hosted service is operated at wan2-7.io.

sh
npm ci --ignore-scriptsnode dist/cli.mjs --helpnode dist/cli.mjs auth loginnode dist/cli.mjs mcp

License and provenance

MIT. See PROVENANCE.md for release provenance. The executable is unchanged from 0.1.1. Version 0.1.3 updates the GitHub repository and MCP Registry namespace to the NotUsingGithub account.

Container for directory introspection

sh
docker build -t wan27-mcp .docker run --rm -i wan27-mcp

The container runs the same pinned client distribution as the npm package, with the required mcp subcommand, as a non-root user. Linux containers explicitly select the private-file credential store because macOS Keychain is unavailable there; the empty container has no credentials. It accepts MCP JSON-RPC on stdin and writes protocol responses on stdout. MCP initialization and tools/list do not require an account or token and do not create media or spend credits. Calling account or generation tools requires an authorized Wan27 account; generation still requires a paid subscription and spends website credits.

The build copies only the distribution and public metadata. No credentials or host configuration are included. Do not embed a token in the Dockerfile or image.

來源:README.md,提交 4b6c08d

工具

0
SourceWeft 尚未在沙箱中啟動過該軟體包,因此其工具未知。

版本歷史

1
  1. v0.1.3最新Oct 12, 2026

評分

已滿足適用檢查項的 50%

該評分衡量此收錄條目的完整程度,依據註冊中心條目、端點與公開來源計算得出。它不是對伺服器本身的評價,也不表示伺服器安全或目前可用:可連通性是帶日期的檢查結果,評分也不包含安裝量或使用者評價。

  • 近 30 天內可連通25 分

    請保持端點能夠回應 MCP 握手;SourceWeft 每 30 天探測一次。

  • 已宣告認證方式10 分

    請在註冊中心條目中宣告憑證請求標頭,或保持端點能夠回應探測,以便判定其認證方式。

  • 已列出工具15 分

    請讓伺服器在無憑證時回應 tools/list,或發布可在沙箱中無憑證啟動的軟體包。

  • 描述必需15 分
  • 圖示10 分
  • 文件15 分
  • 授權條款10 分