ByteDance open-sources UI-TARS-desktop multimodal AI Agent stack; Agent TARS CLI upgraded to v0.3.0
Decision Brief
ByteDance's open-source UI-TARS-desktop is a multimodal AI Agent stack comprising two projects: Agent TARS and UI-TARS Desktop. Agent TARS is a general-purpose multimodal Agent offering a CLI and web UI, with support for hybrid browser agents, event-stream-driven execution, and MCP tool integration. UI-TARS Desktop is a desktop GUI Agent built on UI-TARS and the Seed-1.5-VL/1.6 model series, supporting Windows, macOS, and browsers, with all processing done locally. Agent TARS CLI v0.3.0, released on November 5, 2025, adds streaming support for shell commands and structured multi-file display, runtime settings with tool-call timing stats and deep thinking, an event stream viewer for data-flow tracing and debugging, and exclusive support for the AIO Agent sandbox as an isolated all-in-one tool execution environment. For developers using the command-line Agent for complex tasks, v0.3.0's streaming output makes multi-step execution less of a black box, runtime stats help pinpoint slow tool calls, and sandbox isolation reduces the risk of executing untrusted commands. UI-TARS Desktop's remote computer and remote browser operators are completely free and require no configuration, suiting individual users who need cross-device automation without building infrastructure.
Sources
- Skill Radar(GitHub 趋势)
Trending hands-on MCP servers, agent skills, and AI-coding tools discovered from GitHub search momentum.
- GitHub:bytedance/UI-TARS-desktop
留言
登入后即可留言,和其他 builder 交换实测心得。
还没有留言,抢头香。