Add one URL to Claude, ChatGPT, Cursor or any MCP client and sign in: no install, no key to copy. Ask for an image or a video in plain language and get back a finished file, made with the same models and credits as the studio.
AI Media Studio runs the MCP server for you. Add this URL, sign in once and pick a workspace. You can disconnect any time by revoking the connection under Workspace settings → API.
Server URL
https://app.ai-media-studio.com/mcp
In claude.ai or Claude Desktop: Settings → Connectors → Add custom connector. Paste the server URL above and click Connect.
Sign in to AI Media Studio, pick a workspace and click Allow. The aims tools show up in new chats.
02
Ask in plain language
Mention aims the first time so the agent picks the right tool. It returns hosted URLs and the credits used.
Image
Use aims to make a 16:9 hero image of a red panda astronaut floating above Earth, cinematic light.
Image → video
Now turn that image into a 6-second video with a slow push-in and drifting stars.
Edit your photo
Upload ./product.jpg to aims and put the bottle on wet black slate with soft studio light.
Batch assets
Generate square app icons for these five features with aims and save the URLs to icons.md.
Prefer an API key?
Run the MCP server locally, script the aims CLI or call the REST API. The CLI can also sign in without a key: aims login opens the browser and saves a key for you.
01
Get your API key
In AI Media Studio, open Workspace settings → API and create a key with the scopes your agent needs. Paste it here and every command below is ready to copy.
Your key stays in this browser tab. It is never sent to ai-media-studio.com from this page.
02
Connect your agent
One command adds the aims MCP server for every project.
claude mcp add aims --scope user -e AIMS_API_KEY=aims_... -- npx -y @ai-media-studio/mcp
Or install the plugin (skills + hosted MCP; run /mcp to sign in, no key needed)
claude plugin marketplace add liebhardt-io/aims-agent
claude plugin install aims@aims
Start a new Claude Code session and ask for an image or a video.
What your agent can do
Six tools, the same in the MCP server and the CLI. Every call spends workspace credits only when it generates something; listing models and checking the account are free.
generate_image
Text to image with Nano Banana, GPT Image and more
aims imageedit_image
Change a photo with one instruction
aims editgenerate_video
Text to video and image to video with Kling, Seedance and more
aims videoupload_image
Bring your own photo for edits and image to video
aims uploadlist_models
Live list of models with credit costs
aims modelsget_account
Workspace, credit balance and key scopes
aims whoami
Agent FAQ
What is aims-agent?
aims-agent is the open-source (Apache-2.0) toolkit that connects AI agents to AI Media Studio: the @ai-media-studio/mcp MCP server, the @ai-media-studio/cli command-line tool and agent skills. They call the same public API as the hosted MCP server at app.ai-media-studio.com/mcp, which needs no install.
Which agents can use AI Media Studio?
Any agent that supports MCP: Claude (web and desktop), ChatGPT developer mode, Claude Code, Cursor, Codex, VS Code (Copilot agent mode), Windsurf, Zed and others. Agents without MCP can use the aims CLI or plain HTTP.
Do I need to install anything?
No. Add the hosted server URL to your agent and sign in; there is nothing to install and no key to copy. The local MCP server (npx -y @ai-media-studio/mcp) and the CLI need Node.js 18 or newer and an API key; aims login opens the browser to create one for you.
Which API key scopes does my agent need?
image:generate for images, video:generate for videos and media:write to upload your own photos. Signing in through the hosted server or aims login grants all of them. Only workspace owners and admins can connect agents or create keys, and every connection appears under workspace settings, API, where it can be revoked at any time.
What does it cost?
Agent calls use the same monthly workspace credits as the web app, at the same per-model credit cost. There is no separate API fee; paid plans start at $19/month. Use --dry-run in the CLI to preview a request without spending credits.
Are there rate limits?
Yes, the same on every plan: 60 image and 20 video generation requests per 10 minutes, counted per workspace. Requests over the limit get HTTP 429 with a Retry-After header.
Where do the generated files go?
Every result is saved in your AI Media Studio workspace and returned to the agent as a hosted URL, so you can open it in the app, share it or download it from the agent.