Veo-4-API

概述

Veo-4-API 是 Google DeepMind 的 Veo 4 AI 视频生成 API 的全面 Python 包装器，通过 MuAI 访问。它使开发人员能够创建长达 30 秒的集成音频、角色一致性和高级相机控制的原生 4K 视频。该项目包含专门的 MCP 服务器实现，允许 Claude Desktop 等 AI 助手通过模型上下文协议直接调用视频生成工具。

试试问 AI

装完之后，这里有 6 个你可以让 AI 做的事：

• text_to_video

  Generate native 4K video from text prompt.

  试试:Generate a cinematic tracking shot through a lush rainforest试试:Create a video of a majestic eagle soaring over snow-capped mountains
• image_to_video

  Animate static images into 4K video.

  试试:Animate a landscape photo with clouds drifting and light shifting试试:Bring a beach photo to life with waves crashing and seagulls calling
• text_to_video_with_audio

  Generate video with integrated audio from text prompt.

  试试:Create a video of a busy Tokyo street at night with jazz music试试:Generate a video of a rainforest with birds calling and rain falling
• image_to_video_with_audio

  Animate images with synchronized audio.

  试试:Animate an ocean photo with waves crashing and seagulls calling试试:Bring a static city photo to life with traffic sounds and music
• character_video

  Generate videos with consistent character identity.

  试试:Create a video of this character walking through a neon-lit Tokyo street试试:Show this person exploring an ancient forest with consistent appearance
• extend_video

  Extend an existing video segment.

  试试:Extend the eagle video showing it landing on a mountain peak试试:Continue the beach scene showing the sunset over the ocean
• video_edit

  Edit existing videos with natural language prompts.

  试试:Change the weather to a dramatic thunderstorm in this video试试:Add more vibrant colors to this sunset scene
• upload_file

  Upload local images or videos to the platform.

  试试:Upload this local image of a person for character consistency试试:Upload this video file for editing
• wait_for_completion

  Wait for video generation to complete.

  试试:Check when the video generation is complete试试:Wait for the video processing to finish

可对比工具

安装

pip install veo-4-api

MUAPI_API_KEY=your_muapi_api_key_here

{
  "mcpServers": {
    "veo-4": {
      "command": "python3",
      "args": ["/path/to/Veo-4-API/mcp_server.py"]
    }
  }
}

FAQ

Veo 4 API 是什么？

Veo 4 是 Google DeepMind 最先进的 AI 视频生成模型，具有原生 4K 输出、长达 30 秒的片段、集成音频、角色一致性和高级相机控制功能。

如何获得 Veo 4 的 API 访问权限？

对 Veo 4 的访问通过 MuAI (https://muapi.ai) 提供。您需要从他们的平台获取 API 密钥才能使用此包装器。

支持哪些视频格式和分辨率？

Veo 4 生成原生 4K 视频（非放大），支持 16:9、9:16（抖音/Reels）和 1:1 宽高比，分辨率为 1080p 和 4K 两个质量等级。

Veo-4-API 对比