Give your AI agent35+ image and video models
Connect AI Influencer Studio to Claude, Codex, Cursor or any MCP-compatible agent with our API, MCP server, CLI and Skill. Then just ask: your agent picks the model, writes the prompt, generates the images and videos, adds music and a voiceover, and assembles the final cut.
No card needed · plans from $9.99/month · cancel anytime
35+
Image and video models behind one API key
4 ways
API, MCP server, CLI and Agent Skill
$0.01
Per credit. Listing models and checking credits is free
What you get
Image models
GPT Image 2.5 and 2, Nano Banana 2.1, 2, Lite and Pro, Qwen Image 3, Seedream 5 Lite and 4.5, Krea 2 Turbo, Z-Image Turbo and image editing. Reference images and LoRAs where the model supports them.
Video models
Kling 4.0, Seedance 2 and 2.5, Wan 3 and Wan 3 Prime, Veo 3.1 Fast, Kling 3 Turbo, MiniMax H3, Vidu Q4, P-Video 2, Gemini Omni and LTX 2.3. Text-to-video, image-to-video and reference-to-video.
Full production tools
Suno music, voiceovers and character voices, frame extraction, video assembly, carousel packaging and 4K video upscaling, so your agent delivers finished content, not loose clips.
Your own characters
Agents can list your saved AI influencers and use their character LoRAs. The trigger word and appearance are added automatically, so the face stays the same in every shot.
Works where you work
Claude Code, Claude Desktop, Codex, Cursor, Hermes, OpenClaw and any MCP client. The CLI runs in any terminal or script, and the REST API in your own app.
Cost control built in
Your agent can check your balance and each model's price before generating. Every generation uses app credits at $0.01 each, and API keys can be revoked anytime.
How it works
- 01
Pick a plan
Agent access is included in the Creator and Pro plans, with 3,000 or 12,000 monthly credits.
- 02
Create an API key
Open Agent access in the app and create a key. It is shown once, and you can revoke it anytime.
- 03
Connect your agent
Paste the MCP config into Claude, Codex or Cursor, install the Skill, or set the key in the CLI.
- 04
Just ask
"Make a 30-second ad for my sneakers with music and a voiceover." Your agent does the rest.
Generate anything, not only AI influencers
The same models that power the studios are available to your agent, so you can produce whatever you need: product ads and UGC videos, short films and trailers, music videos, faceless YouTube channels, explainer videos, Instagram carousels, product photos, thumbnails and, of course, AI influencer content.
A typical run: your agent writes a shot list, generates a key image for each scene with GPT Image 2.5 or Nano Banana, animates each one with Kling 4.0, Seedance 2.5 or Wan 3, creates a Suno soundtrack and a voiceover, then assembles the final video. You review the result instead of switching between ten tabs.
Connect Claude, Codex or Cursor with MCP
MCP (Model Context Protocol) lets an AI agent use outside tools. Add AI Influencer Studio to your agent's MCP config with your API key, restart it, and the tools appear: list_influencers, generate_image, generate_video, generate_music, generate_voiceover, assemble_video, upscale_video and more.
The Agent Skill teaches your agent which model fits which job, the right settings for each one, credit costs and how to recover from errors, so results are good from the first try.
{
"mcpServers": {
"ai-influencer-studio": {
"command": "npx",
"args": [
"-y", "mcp-remote@latest",
"https://www.aiinfluencerstudio.app/api/mcp",
"--header", "Authorization: Bearer ais_your_api_key_here"
]
}
}
}Prefer scripts? Use the CLI
The aiis command-line tool gives you the same models in any terminal, script or automation. Generate an image of your influencer, animate it, or assemble a video from a JSON edit list, then plug it into your own pipeline.
aiis auth set-key ais_your_api_key_here
aiis generate image --influencer Naomi --prompt "cinematic luxury hotel lobby portrait"
aiis generate video --model veo-3.1-fast-i2v --ref ./portrait.png --prompt "cinematic fashion reel" --duration 8
aiis tool assemble_video --json-file ./edit.jsonQuestions, answered
What is an MCP server for AI image and video generation?
It's a connector that lets an AI agent like Claude, Codex or Cursor call image and video models directly. Once connected, you ask in plain words and the agent generates, checks and assembles the media for you.
Which AI agents and apps work with it?
Claude Code, Claude Desktop, Codex, Cursor, Hermes, OpenClaw and any client that supports MCP. You can also use the aiis CLI in a terminal or call the REST API from your own app.
Can I generate things other than AI influencers?
Yes. The image and video models are general purpose: ads, product videos, short films, music videos, YouTube content, thumbnails, carousels and more. Saved influencers and character LoRAs are an extra, not a requirement.
Which models are included?
More than 35 image and video models, including Kling 4.0, Seedance 2 and 2.5, Wan 3, Veo 3.1 Fast, Kling 3 Turbo, MiniMax H3, P-Video 2, GPT Image 2.5, Nano Banana 2.1 and Pro, Qwen Image 3, Seedream and Krea 2 Turbo. New models are added as they launch.
How much does it cost?
Agent access is included in the Creator ($24.99/month, 3,000 credits) and Pro ($79.99/month, 12,000 credits) plans. Each generation uses credits at $0.01 per credit, the same price as in the app. Listing your influencers and checking your balance are free.
What is the difference between the API, MCP, CLI and Skill?
The MCP server plugs the tools into AI agents. The Skill teaches the agent how to use them well. The CLI runs them from a terminal or script, and the REST API lets you call them from your own code. One API key works for all four.
Is my API key safe?
Keys are shown once when you create them and can be revoked anytime from the Agent access page. Every generation is charged to your own credit balance, which your agent can check before big batches.
Connect your AI agent today
Every new account gets 100 free credits. No card needed.
Get 100 free credits