TinyClick: Single-Turn Agent for Empowering GUI Automation
Pawel Pawlowski, Krystian Zawistowski, Wojciech Lapacz, Adam Wiacek, Marcin Skorupa, Sebastien Postansque, Jakub Hoscilowicz
- 🏛 Institutions
- Samsung R&D Poland, Warsaw University of Technology
- 📅 Date
- October 9, 2024
- 📑 Publisher
- INTERSPEECH 2025
- 💻 Env
- Desktop Mobile Web
- 🔑 Keywords
TLDR
TinyClick is a 0.27B single-turn GUI agent built on Florence-2-Base that predicts the target UI element from a screenshot and user command. The paper attributes its gains to vision-specific multitask training and MLLM-based data augmentation, and reports strong results on ScreenSpot and OmniAct annotations while keeping latency and training cost low.
Related papers
- Ponder & Press: Advancing Visual GUI Agent towards General Computer ControlDecember 2, 2024 · Findings of ACL 2025
- SeeClick: Harnessing GUI Grounding for Advanced Visual GUI AgentsJanuary 17, 2024 · ACL 2024
- VenusBench-GD: A Comprehensive Multi-Platform GUI Benchmark for Diverse Grounding TasksDecember 18, 2025 · arXiv
- GuirlVG: Incentivize GUI Visual Grounding via Empirical Exploration on Reinforcement LearningAugust 6, 2025 · ICLR 2026 (Poster)
- ScaleTrack: Scaling and back-tracking Automated GUI AgentsMay 1, 2025 · arXiv
- UI-TARS: Pioneering Automated GUI Interaction with Native AgentsJanuary 21, 2025 · arXiv