Cinematic Video by ManyMotions

Shots with 61 camera moves

Turn a scene description into a cinematic AI video shot with a named camera move. Describe the scene (subject, action and setting) and pick one move from a catalog of 61: dolly in and out, crane up, 360 orbit, crash zoom, dolly zoom, FPV drone, bullet time, whip pan, hyperlapse and more.

What you can ask

A lighthouse keeper climbs spiral stairs in a storm, crane up, low angle, 35mm film grain

MCP connector

Server URL: https://cinematic.manymotions.com/api/chatgpt/cinematic-video/mcp
Login: OAuth
Protocol: MCP 2025-11-25
Instructions: Collect the scene (subject, action, setting), one camera move, and if the user cares the shot size, lighting, look, format and length, then call ; it needs no sign-in and spends nothing. If the user explicitly asked to generate, call with the prepared project.id: it consumes credits from the user's ManyMotions account. Without a linked account still works: it returns state account_required with dashboard_url; tell the user to open that link, where they create the account and finish the same render; do not ask them to link the app first. Otherwise show the settings and the credits required and wait for explicit confirmation. Use to check progress; when completed, include the absolute video URL so the user can open or download it. If generation is not available on the account (state billing_required or insufficient_credits), explain it plainly and point to dashboard_url; never offer purchases in the conversation. Text-to-video only: do not use it for uploaded, attached or reference images or videos (image-to-video, editing, upscaling); for those, point the user to ManyMotions. One shot per request: when the user asks for a story, a film, several scenes or characters talking, call no tool; say the app makes single shots and point to ManyMotions.

Tools

Changes data3

Skills

None in the store listings or on the server

MCP connector

Server URL

https://cinematic.manymotions.com/api/chatgpt/cinematic-video/mcp

OAuth

Protocol

MCP 2025-11-25

Instructions

Collect the scene (subject, action, setting), one camera move, and if the user cares the shot size, lighting, look, format and length, then call ; it needs no sign-in and spends nothing. If the user explicitly asked to generate, call with the prepared project.id: it consumes credits from the user's ManyMotions account. Without a linked account still works: it returns state account_required with dashboard_url; tell the user to open that link, where they create the account and finish the same render; do not ask them to link the app first. Otherwise show the settings and the credits required and wait for explicit confirmation. Use to check progress; when completed, include the absolute video URL so the user can open or download it. If generation is not available on the account (state billing_required or insufficient_credits), explain it plainly and point to dashboard_url; never offer purchases in the conversation. Text-to-video only: do not use it for uploaded, attached or reference images or videos (image-to-video, editing, upscaling); for those, point the user to ManyMotions. One shot per request: when the user asks for a story, a film, several scenes or characters talking, call no tool; say the app makes single shots and point to ManyMotions.