Surfer 2: The Next Generation of Cross-Platform Computer Use Agents
Andreux, Mathieu, Bakler, Märt, Barbier, Yanael, Benchekroun, Hamza, Biré, Emilien, Bonnet, Antoine, Bordie, Riaz, Bout, Nathan, Brunel, Matthias, Cambray, Aleix, Cedoz, Pierre-Louis, Chassang, Antoine, Cloix, Gautier, Connelly, Ethan, Constantinou, Alexandra, De Coster, Ramzi, de la Jonquiere, Hubert, Delfosse, Aurélien, Delpit, Maxime, Deprez, Alexis, Derupti, Augustin, Diaz, Mathieu, D'Souza, Shannon, Dujardin, Julie, Edmund, Abai, Eickenberg, Michael, Fatalot, Armand, Felissi, Wissem, Herring, Isaac, Koegler, Xavier, de Kergaradec, Erwan Le Jumeau, Lac, Aurélien, Langevin, Maxime, Lauverjat, Corentin, Loison, Antonio, Manevich, Avshalom, Moyal, Axel, Kerbel, Axel Nguyen, Parovic, Marinela, Revelle, Julien, Richard, Guillaume, Richter, Mats, Riochet, Ronan, Santos, María, Savidan, Romain, Sifre, Laurent, Theillard, Maxime, Thibault, Marc, Valentini, Ivan, Wu, Tony, Yie, Laura, Yuan, Kai, Zubovskij, Jevgenij
–arXiv.org Artificial Intelligence
Building agents that generalize across web, desktop, and mobile environments remains an open challenge, as prior systems rely on environment-specific interfaces that limit cross-platform deployment. We introduce Surfer 2, a unified architecture operating purely from visual observations that achieves state-of-the-art performance across all three environments. Surfer 2 integrates hierarchical context management, decoupled planning and execution, and self-verification with adaptive recovery, enabling reliable operation over long task horizons. Our system achieves 97.1% accuracy on WebVoyager, 69.6% on WebArena, 60.1% on OSWorld, and 87.1% on AndroidWorld, outperforming all prior systems without task-specific fine-tuning. With multiple attempts, Surfer 2 exceeds human performance on all benchmarks. These results demonstrate that systematic orchestration amplifies foundation model capabilities and enables general-purpose computer control through visual interaction alone, while calling for a next-generation vision language model to achieve Pareto-optimal cost-efficiency.
arXiv.org Artificial Intelligence
Oct-27-2025
- Genre:
- Research Report > New Finding (0.48)
- Industry:
- Information Technology > Services (0.46)
- Technology: