minimax-video-genlisted

# MiniMax Video Generation Skill > **What this skill does — generates videos.** > Given a text prompt (or a prompt + reference image / first-frame / last-frame), > this skill calls the MiniMax video model and returns an actual video file on > disk. It is the skill to load whenever the user asks for "做个视频", "出一段视频", > "拍一段短片", "生成一个动画", "做一段 demo 视频", "make a video", "create a short", > "generate an animation", "render a clip", or any moving-picture deliverable. The skill wraps the official MiniMax endpoints: ``` POST {API_BASE_URL}/v1/video_generation # async create (legacy models) POST {API_BASE_URL}/v2/video_generation # multimodal create (MiniMax-H3) GET {API_BASE_URL}/v1/query/video_generation # poll status GET {API_BASE_URL}/v1/files/retrieve # resolve download URL ``` It ships: - a reusable Python client (`scripts/video_client.py`), - parameter validators (`scripts/validators.py`), - a CLI entry point (`scripts/generate_video.py`) — the **primary** interface, invoked via `bash` when the skill is in scope. ## Capabilities (when to use this skill) Use it whenever the user wants to **generate a video**, including: - **Text-to-video (文生视频 / t2va)** — turn a written description into a clip. - **Image-to-video (i2va)** — first-frame image + text, "make this picture move". - **First + last frame (start/end to video)** — define both ends and let the model interpolate. - **Reference-to-video (r2va)** — keep a character / scen