Agnes Video MCP Server

MCP server for Agnes Video v2.0 (text-to-video / image-to-video generation), via OpenAI-compatible API.

Features

• Text-to-Video: Generate videos from text prompts
• Image-to-Video: Generate videos from a starting image or keyframes
• Async workflow: Submit task → poll for result
• Custom resolution, frame count, frame rate, duration, seed

Requirements

• Node.js >= 18
• Agnes AI API key (free at https://platform.agnes-ai.com)

Environment Variables

Variable	Required	Default	Description
AGNES_API_KEY	✅	—	Agnes AI API key
AGNES_VIDEO_MODEL		agnes-video-v2.0	Model name

Tools

agnes_video_create — Submit a video generation task

Parameter	Type	Required	Default	Description
prompt	string	✅	—	Text description
width	number		1152	Video width
height	number		768	Video height
num_frames	number		121	Frames (8n+1: 121, 241, 361, 441)
frame_rate	number		24	Frame rate
duration	number			Duration in seconds
seed	number			Reproducible seed
image_url	string			Starting image URL
mode	string			ti2vid or keyframes

agnes_video_poll — Check task progress

Parameter	Type	Required	Description
task_id	string	✅	Task ID returned from agnes_video_create

Usage

1. agnes_video_create → "a cat walking on beach at sunset"
   → Returns task_id: "task_xxx"

2. agnes_video_poll → task_id: "task_xxx"
   → Returns status + progress %, or video URL when done

License

MIT