🔥 News

• 2026.05: 🔥 Released ODE, a closed-loop data evolution framework for visual-native multimodal deep search agents, enabling reusable visual evidence and policy-aware SFT/RL data curation.
• 2026.03: 🎉 One paper (AdaCtrl) is accepted by TMLR.
• 2026.02: 🎉 We released Seed 2.0! It is a powerful unified model with comprehensive capabilities, featuring exceptionally advanced agent capabilities. Check out the model card for more details!
• 2026.01: 🎉 Three papers are accepted by ICLR 2026.
• 2025.12: 🎉 Seed 1.8 is live! Welcome to check the model card for more details!
• 2025.09: 🎉: We’re excited to release UI-TARS-2, an all-in-one agent model with stronger GUI, game, code, and tool- use capabilities. It integrates multiple abilities for complex tasks.
• 2025.04: 🔥 Released ReTool, a reinforcement learning-based multi-turn tool-use agent training framework!
• 2025.01: 🎉 Released UI-TARS: Industry’s open-source GUI Agent foundation model with 6.2K+ GitHub stars!
• 2024.12: 🎉 One paper (SKE-Learn) is accepted by AAAI 2025 (Oral).
• 2024.05: 🎉 One paper (UltraTool) is accepted by ACL 2024 (Findings).