Create an API key, add this server configuration, then ask your agent to create a video.
In claude.ai or the Claude desktop app, go to Settings > Connectors > Add custom connector, name it ViewMax, paste this URL, then click Connect and sign in with your ViewMax account. No API key needed.
https://viewmax.studio/api/mcp
Or paste one sentence to your agent
Agents that can edit their own MCP configuration can connect themselves. Replace YOUR_API_KEY, then paste:
Add the ViewMax MCP server to this client: streamable HTTP transport, URL https://viewmax.studio/api/mcp, HTTP header "Authorization: Bearer YOUR_API_KEY". Then call the list_video_models tool to verify the connection.
MCP for Claude
How to add ViewMax MCP to Claude
Step-by-step setup for Claude Desktop, claude.ai, and Claude Code. Both connect to the same server: https://viewmax.studio/api/mcp
Claude Desktop & claude.ai
Open Settings, then Connectors, and click "Add custom connector".
Name the connector ViewMax and paste the server URL: https://viewmax.studio/api/mcp
Click Connect and sign in with your ViewMax account. OAuth handles authentication, so no API key is needed.
Start a new chat and ask Claude to generate a video, image, music, or speech. Approve the ViewMax tools when Claude asks.
Claude Code
Create an API key in Settings > API Keys on ViewMax.
Or commit this config as .mcp.json in the project root. Claude Code reads the key from each developer's VIEWMAX_API_KEY environment variable, so no key is committed:
The ViewMax MCP server generates video, image, music, and audio. Flagship images on Pro/Ultra can use the shared daily fair-use pool; other generations consume credits.
generate_video
Create a video task; discover options with list_video_models and get_video_model.
generate_image
Text-to-image or image-to-image; browse models with list_image_models.
generate_music
Create an original music track from a prompt; see list_music_models.
generate_speech
Convert text to speech with a chosen voice from list_voices.
generate_sound_effect
Generate a sound effect from a text prompt.
get_task
Get a task's current status and completed output URLs.
wait_for_task
Wait up to 50 seconds, then return the latest status. Call it repeatedly for longer generations.
get_credits
Check the remaining credits in your authenticated account.
Example
From an instruction to a video URL
You
Make a 10s 720p video of a paper kite crossing a bright summer sky.
Your agent calls generate_video
{ "model": "gemini-omni", "prompt": "A paper kite crossing a bright summer sky", "duration": "10s", "resolution": "720p" }