inclusionAI/UI-Venus

Health v2 · maintenance only

Flags: no_releases no_license

Adoption not part of the score

1032 stars · 84 forks observed · 2026-09-03

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

UI-Venus is a family of open-source multimodal GUI agent models (9B/27B) that perform UI element grounding and task navigation from screenshots alone across mobile, web, and desktop environments. It is trained with reinforcement learning using keypoint-grounded verification and reflection supervision.

Use cases

• automate tasks in mobile apps from screenshots
• ground and click UI elements on a screen
• build a computer-use agent for desktop OS interaction
• navigate web pages autonomously with a vision-language model
• benchmark GUI agents on mobile tasks
• train a UI agent with reinforcement learning

When to choose

• you need screenshot-only GUI automation across mobile, web, and desktop
• you want an open-weights foundation GUI agent to fine-tune or deploy
• you need robust long-horizon task completion with reflection and verification

When to avoid

• you need DOM/accessibility-tree-based automation rather than vision-only grounding
• you need a lightweight non-ML automation tool like Selenium or Playwright
• you lack GPU resources to run 9B/27B models

Facets

library · maturity active

agent-framework computer-vision machine-learning reinforcement-learning llm-inference artificial-intelligence computer-vision python cross-platform gui-agent ui-grounding multimodal-llm screenshot-input computer-use mobile-automation huggingface-model ai-agents automation gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="inclusionAI/UI-Venus")