MCP server for Longtake Studio — let an AI agent shoot AI video (Seedance, Veo, Gemini Omni, Kling) on your own API key
Setup
Choose your client, copy the config, then fill in any required environment variables.
claude mcp add longtake --env STUDIO_URL='<STUDIO_URL>' --env STUDIO_TOKEN='<STUDIO_TOKEN>' -- npx -y longtake-mcp~/.codex/config.toml
[mcp_servers.longtake]
command = "npx"
args = ["-y", "longtake-mcp"]
env = { STUDIO_URL = "<STUDIO_URL>", STUDIO_TOKEN = "<STUDIO_TOKEN>" }
# codex mcp add longtake --env STUDIO_URL='<STUDIO_URL>' --env STUDIO_TOKEN='<STUDIO_TOKEN>' -- npx -y longtake-mcp~/.cursor/mcp.json
{
"mcpServers": {
"longtake": {
"command": "npx",
"args": [
"-y",
"longtake-mcp"
],
"env": {
"STUDIO_URL": "<STUDIO_URL>",
"STUDIO_TOKEN": "<STUDIO_TOKEN>"
}
}
}
}.vscode/mcp.json
{
"servers": {
"longtake": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"longtake-mcp"
],
"env": {
"STUDIO_URL": "<STUDIO_URL>",
"STUDIO_TOKEN": "<STUDIO_TOKEN>"
}
}
}
}~/.gemini/settings.json
{
"mcpServers": {
"longtake": {
"command": "npx",
"args": [
"-y",
"longtake-mcp"
],
"env": {
"STUDIO_URL": "<STUDIO_URL>",
"STUDIO_TOKEN": "<STUDIO_TOKEN>"
}
}
}
}claude_desktop_config.json
{
"mcpServers": {
"longtake": {
"command": "npx",
"args": [
"-y",
"longtake-mcp"
],
"env": {
"STUDIO_URL": "<STUDIO_URL>",
"STUDIO_TOKEN": "<STUDIO_TOKEN>"
}
}
}
}.trae/mcp.json
{
"mcpServers": {
"longtake": {
"command": "npx",
"args": [
"-y",
"longtake-mcp"
],
"env": {
"STUDIO_URL": "<STUDIO_URL>",
"STUDIO_TOKEN": "<STUDIO_TOKEN>"
}
}
}
}Environment variables
STUDIO_URLSTUDIO_TOKEN
Replace the placeholders with your own credentials and keep them out of version control.