por bytedance · bytedance/UI-TARS-desktop
ByteDance's open multimodal agent stack — vision-grounded computer-use, callable from Claude via MCP.
UI-TARS Desktop is ByteDance's open multimodal agent platform. It uses screen understanding for desktop control and exposes itself as MCP. Claude can delegate vision-grounded GUI tasks where text-only tools fall short.
%APPDATA%\Claude\claude_desktop_config.json{
"mcpServers": {
"ui-tars-desktop-mcp": {
"command": "docker",
"args": [
"run",
"-p",
"8080:8080",
"ghcr.io/bytedance/ui-tars-desktop:latest"
]
}
}
}Abre Claude Desktop → Settings → Developer → Edit Config. Reinicia después de guardar.
{
"mcpServers": {
"ui-tars-desktop-mcp": {
"command": "docker",
"args": [
"run",
"-p",
"8080:8080",
"ghcr.io/bytedance/ui-tars-desktop:latest"
]
}
}
}Cursor usa el mismo esquema mcpServers que Claude Desktop. La configuración del proyecto prevalece sobre la global.
{
"mcpServers": {
"ui-tars-desktop-mcp": {
"command": "docker",
"args": [
"run",
"-p",
"8080:8080",
"ghcr.io/bytedance/ui-tars-desktop:latest"
]
}
}
}Haz clic en el icono MCP Servers de la barra lateral de Cline y luego en "Edit Configuration".
{
"mcpServers": {
"ui-tars-desktop-mcp": {
"command": "docker",
"args": [
"run",
"-p",
"8080:8080",
"ghcr.io/bytedance/ui-tars-desktop:latest"
]
}
}
}Mismo formato que Claude Desktop. Reinicia Windsurf para aplicar.
{
"mcpServers": [
{
"name": "ui-tars-desktop-mcp",
"command": "docker",
"args": [
"run",
"-p",
"8080:8080",
"ghcr.io/bytedance/ui-tars-desktop:latest"
]
}
]
}Continue usa un array de objetos de servidor en lugar de un mapa.
{
"context_servers": {
"ui-tars-desktop-mcp": {
"command": {
"path": "docker",
"args": [
"run",
"-p",
"8080:8080",
"ghcr.io/bytedance/ui-tars-desktop:latest"
]
}
}
}
}Añádelo a context_servers. Zed recarga en caliente al guardar.
claude mcp add ui-tars-desktop-mcp -- docker run -p 8080:8080 ghcr.io/bytedance/ui-tars-desktop:latestUn solo comando. Verifica con claude mcp list. Quita con claude mcp remove.
Cuándo usarlo: You're automating a desktop or canvas app that has no scripting surface.
Resultado: GUI automation in apps with no API.
UI-TARS for vision tasks; OpenAgent for code-text tasks
| Herramienta | Entradas | Cuándo llamar | Coste |
|---|---|---|---|
| ui_act | (see docs) | Run a vision-grounded GUI action | 1 call |
| ui_observe | (see docs) | Capture screen state | 1 call |
screen-capture input-injectionRun with banking app visibleAction misses target on Retina displaySet the DPI scale; UI-TARS defaults to 1x
| Alternativa | Cuándo usarla | Contrapartida |
|---|---|---|
| OpenAgent | Lower resource cost | Less vision grounding |