Agent S4 Evolves Its Own Computer-Use Interfaces
Evolving Agent Interfaces
Interface design itself becomes a learnable, evolving capability, meaning smaller open models can beat frontier systems on computer-use tasks at a fraction of the cost. Builders working on agents can reuse the open-source repository and apply the evolution loop to their own workflows. The results set a concrete cost-performance benchmark against GPT-5.5 and Claude Opus 4.8 that teams choosing an agent stack can weigh directly.
Agent S4 is an autonomous computer-use agent that builds and revises Evolving Agent Interfaces from task experience. A meta-agent analyses task-agent trajectories on synthetic training tasks without success labels, then uses population-based evolution to select repositories that generalise on unseen validation tasks. On OSWorld V2, paired with Qwen3.8-27B, it reaches a 21.3% binary score, 3.8 times the same model with fixed interfaces, above GPT-5.5 (13.0%) and Claude Opus 4.8 (20.6%) at about 4.3% and 1.5% of their costs. On ScienceBoard, evolved interfaces lift success from 41.4% to 58.6% with the same model. The code is open-sourced on GitHub.
Evolving Agent Interfaces
Computer-use agents are limited by the interfaces humans give them. An agent may type a spreadsheet cell by cell when a single line of code would do, or print all two thousand lines of a file through a shell to find one function. Both the interfaces and the knowledge of when to use each are built by hand, which cannot keep pace with new applications, workflows, and models. We argue that interface design should itself be an agent capability. We introduce Agent S4, an autonomous computer-use scientist that builds Evolving Agent Interfaces from task experience. A meta-agent examines task agents' trajectories on synthetic training tasks, without success labels, and constructs and revises interfaces and usage knowledge in a repository. Population-based evolution then selects which repositories to develop further by their performance on validation tasks the meta-agent never sees, favoring changes that generalize beyond the seen tasks. On OSWorld V2, Agent S4, with Qwen3.8-27B and the evolved repository of interfaces, sets a new state of the art in open-source computer-use agents, reaching 21.3% binary score, 3.8 times the same model with fixed interfaces. This is above both GPT-5.5 (13.0%) and Claude Opus 4.8 (20.6%), at about 4.3% and 1.5% of their respective costs. On ScienceBoard, the evolved interfaces improve the success rate from 41.4% to 58.6% with the same model. These results suggest that useful interfaces can be evolved from experience, making autonomous interface design a source of improvement complementary to advances in models and harnesses. We open-source our code at https://github.com/simular-ai/agent-s.
