No API Key. Simple Setup.
Set up Pollo MCP or CLI, then sign in to your Pollo account through the browser. No API credentials to create, copy, or manage.
Pollo MCP for AI Agents
Connect Pollo MCP to Claude, ChatGPT, or Cursor to create images and videos from prompts, files, or references inside your agent conversation. Keep your creative workflow in one place without switching tools.
1
Copy the Pollo MCP URL
Copy the Pollo MCP server URL. You'll need it in the next step.
https://mcp.pollo.ai/mcp
2
Add Pollo to Claude
Open Settings → Connectors → Add custom connector. Name it Pollo and paste the URL.
3
Connect and Start Creating
Click Add → Connect, sign in with your Pollo account, then ask Claude to create an image or video.
Create With Pollo AI
Create Images and Videos from Text, Images, or References
Bring a rough idea, a product shot, or a supported reference you already trust. Turn what is already in front of you into a polished image or video, without waiting for a perfect starting point.
Build From Your Agent's Existing Context
Need a hero image halfway through a landing page or a game asset while coding? Ask for it right there, using the context your agent already has instead of briefing another tool from scratch.
Keep Products, Characters, and Styles Consistent
Once a character, product, or visual direction feels right, do not lose it in the next generation. Use supported references to keep new images and videos closer to the approved look, including when you turn a still into motion.
Let Your Agent Choose the Right Model and Settings
You should not need to study model names and parameters just to make a cinematic reveal, social ad, or illustrated scene. Tell your agent the outcome, and it can find compatible models and valid settings or follow the model you choose.
Monitor Your Budget and Track Generation Status
Check your available credits and estimate the cost before generating, so you can stay in control of your budget. Once submitted, your agent can track the task status and return the finished image or video when it is ready.
Integrate Finished Assets Into Your Workflow
Start the generation and keep working. When the image or video is ready, your agent brings it back as material for the next step, whether you are building a game, publishing a page, or creating the next asset.
The Latest Models, All in One Place
Give your agent access to a broad catalog of current image and video models through Pollo MCP & CLI.
Why Connect Pollo AI to Your Agent
Images and Videos, One Creative Flow
Create images and videos from text or your own assets in one connected flow. Use a generated image as the starting point for another image or a video.
Latest Models, One Connection
Create with the latest versions of leading models, including Seedance 2.5, Kling 3, Veo 3.1, and GPT Image 2, through one Pollo AI connection.
More References, Better Consistency
Upload multiple images of your product, character, or style to give supported models more context and keep results closer to your intended look.
No Tool Switching After Generation
Get finished images and videos back through your agent, ready to review, download, or use in the next step, so you can keep building in the same workflow.
Clear Costs, One Balance
See the credit estimate before each generation and use the Pollo balance already in your account, making larger creative runs easier to budget.
Frequently Asked Questions
How does Pollo connect to AI agents?
Pollo connects to AI agents through MCP or CLI: MCP works like a plug-in inside compatible AI apps, while CLI lets coding agents use Pollo through terminal commands. Just describe what you need in natural language, and your agent handles the technical work.
Which AI agents are supported?
The initial release supports Claude, ChatGPT, Codex, and Cursor. OpenClaw and Hermes integrations are coming soon.
What can I create?
You can create images and videos from text, existing images, or supported references. Available inputs, formats, durations, and resolutions depend on the selected model.
Can I choose the model myself?
Yes. Specify a model directly or ask your agent to compare available models and recommend one for your brief, format, and budget.
Can I use reference images, videos, or audio?
Yes, with compatible models. Supported reference types and limits vary, so your agent checks the model fields before generation.
Can I upload images from my computer?
Yes. Remote MCP provides an interactive upload experience that turns local files into secure URLs the generation tools can use.
What is the best MCP server for AI image and video generation?
Pollo MCP is the best all-in-one MCP server for AI image and video generation. It brings leading models such as Seedance, Kling, GPT Image, and Nano Banana into one workflow, with cost estimates, model selection, and progress tracking built in.
Do I need an API key?
No. Remote MCP uses browser-based Pollo account authorization, while CLI and local MCP reuse your Pollo CLI login.
How does pricing work?
MCP and CLI use your existing Pollo credits. Your agent can check your available balance and estimate the expected cost before submitting a generation.
How do I receive the finished result?
On MCP Apps hosts, the live generation card updates in place. On text-only clients, your agent checks the task status and returns the finished URLs.
Connect Once.
Keep Creating.
Bring Pollo into your agent's workflow and turn ideas, images, and references into ready-to-use visuals.