Seevio MCP Guide: Generate AI Images and Videos

1. Connect to Seevio MCP

Connect Seevio to create videos and images from your AI conversation, without installing a video model on your computer.

Have your Seevio account ready and choose a client with remote MCP (Streamable HTTP) and OAuth support. You do not need an API key. Before generating, make sure your selected account has enough credits.

1. In your AI client, add a remote MCP server named Seevio with the address shown above. Choose OAuth or automatic authentication.
2. Follow the sign-in link, log in or create your Seevio account, choose personal or available team credits, and approve media service access. This covers current and future models.
3. Return to the AI client and enable Seevio in the conversation. If your client asks to restart or reconnect, follow that prompt.
4. Ask: “List the video and image models available through Seevio.” You can connect even if your Seevio account has never had an API key.

2. Choose your AI client

Find your AI client below and follow its steps. If a menu is missing, use the official guide linked in the card or ask your workspace administrator for access.

codex mcp add seevio --url 'https://api.seevio.ai/mcp'
codex mcp login seevio
codex mcp list

claude mcp add --transport http seevio 'https://api.seevio.ai/mcp'
claude mcp list

# In Claude Code: /mcp → seevio → Authenticate

3. What you can do today

Use Seedance for videos or the image models below for text-to-image and image-to-image. Tell your assistant what you want to create; it can check the model settings and prepare a quote first.

• Generate a video from a text description.
• Animate an image or use images, video and audio as references where the chosen model supports them.
• Text-to-image and image-to-image: Nano Banana Pro, Nano Banana 2, Nano Banana, GPT Image 2.5, and GPT Image 2. Each task generates one image.
• Prepare a request and read its credit quote before generation.
• Create a Seedance 2.5 Draft and separately generate a 1080p final from a completed, eligible Draft.
• Check generation progress, retrieve result links and list tasks created by the current connection.
• Read the selected account balance and permissions when you grant usage access.

4. Use your own reference files

Attach your image, video or audio file in the AI chat and ask: “Use this reference to generate a video or image with Seevio.” If your client supports attachment transfer, your assistant imports it and handles the asset ID for you.

If your assistant cannot read the attachment, provide a downloadable HTTPS file URL, or upload it in connection management and send the returned asset ID to your assistant. Clients with file access can also use the direct-upload tools.

Check the file size before uploading: images up to 20 MB, videos up to 100 MB and audio up to 15 MB. Your selected model may also have format, duration and reference-count requirements; ask your assistant to check them before generating. Nano Banana accepts reference images up to 10 MB each.

5. Credits, permissions and authorization

Connecting, browsing models, checking progress and preparing a quote do not cost generation credits. When you submit a generation, we reserve credits from your selected Seevio account. Your AI client subscription and Seevio credits are billed separately.

Your connection stays authorized without per-task or cumulative credit caps. Keep enough credits in your account, and revoke access whenever you want in connection management. Already accepted tasks may still complete and settle. After a long time away, your client may ask you to sign in again.

A Draft and its final video are separate paid tasks. Ask your assistant for each quote before submitting. With automatic duration, we reserve an upper amount and return unused credits after settlement. If a task is being reconciled, ask your assistant to check that task instead of generating another.

6. Try these requests

Mention Seevio explicitly when you have several generation services connected. Use the same connection when checking a previous task.

7. What is planned next

We plan to expand MCP so you can do more of your creative work in the same AI conversation. These features are planned next:

• More video models: choose the right model for each project using the connection and generation flow you already know.
• More image models for different visual styles and editing needs.

8. Common questions

9. When do you need an API key, and how do you enable it?

API keys are an alternative for clients without OAuth support and automated programs. If you already connected with OAuth, you do not need to create or enable an API key.

API keys and OAuth provide the same MCP features. MCP authorization is managed separately from the key’s REST API permissions and covers current and future media models. Enabling the same key again replaces its previous MCP authorization.

1. Use the API Key link below to create a key, or use an existing one.
2. Open connection management and expand “Advanced connection methods.” Choose your API key and personal or team billing account. Accept the authorization and select “Enable an API key for MCP.” The connection will allow paid video and image generation.
3. In your client or automated program, enter the MCP URL shown at the top of this page. Store the key in secure credential settings and choose Bearer Token authentication; for request headers, use Authorization: Bearer <API_KEY>. Do not paste the key into AI chats or shared configuration files.
4. Check the available models to verify the connection. You can revoke the connection at any time in connection management.