MCP guides
How to generate AI video inside Claude, without leaving the chat
Most people who want AI video bounce between five tools. One for images, one for video, one for voice, one for music, one to stitch it together. Then they pay for all five. StoryStudio takes a different route: you stay in your agent, and the studio works for you from the chat.
Here is how it works.
What StoryStudio actually is
It is an MCP server. You connect it to Claude, Cursor, Codex or any agent that speaks MCP, and you ask for what you want in plain language. Claude writes the story, the studio generates the shots, and you get the result back without switching apps.
Under the hood it pulls the best models on the market into one place: Nano Banana 2, GPT-Image 2, FLUX, Veo 3.1, Seedance 2.0, Kling and more. You do not pick a model and get stuck with it. You get the right one for the shot.
Connect it once
Open your agent's connector settings, add the StoryStudio MCP server, paste the server URL, and you are set. No separate keys to juggle for each model, no extra video app to open. The connection holds, so the next time you just ask.
Describe what you want
Talk to it the way you would brief a team. "A beauty series for my launch, five episodes, vertical, in three languages." Or smaller: "Turn this still image into a ten second cinematic clip." The agent reads the brief and runs the generation.
Because StoryStudio hosts your reference images itself, you can drop a photo of your product or your brand into the chat and it gets used as a reference. No third party file host, no extra step.
Keep your characters consistent
This is where most AI video falls apart. Shot two and the face has changed. StoryStudio keeps a story bible with your cast and your world, so a character stays the same person across every shot. That is the difference between a string of clips and an actual scene.
Edit and export in the same flow
Generation is half the job. StoryStudio adds a timeline editor in the browser, voice-over and music, and an MP4 export. You go from idea to a finished vertical video in one conversation, ready to publish.
Why this beats stitching tools together
Running raw generation across separate apps is slow and the pieces never match. Keeping it in your agent, with one set of models and one editor, is faster and the result holds together. The MCP space is still young, so this is a fresh way to work, not a crowded one.
FAQ
How do I generate AI video directly in Claude? Connect the StoryStudio MCP server to Claude, then describe your video in plain language. Claude writes the story and the studio generates, edits and exports it.
Do I need an API key for each model? No. StoryStudio gives you access to many models through one connection, so you do not manage separate keys.
Can it keep the same character across shots? Yes. The story bible holds your cast and world, so characters stay consistent from one shot to the next.
Want to generate and edit AI video from your agent? Get started with StoryStudio.