SynthAPI Pipeline Derives Scaling Laws for Tool-Use Agents
Scaling Laws for Tool-Use Agents via Controllable Environment Synthesis
The non-obvious finding is that data composition, not raw task count, drives tool-use agent performance, giving teams a concrete recipe for allocating synthetic data budgets. Builders training or fine-tuning agents can apply SynthAPI's control axes to prioritize tool and database diversity over volume. Compared with prior synthetic environment generation work, this approach offers a quantitative scaling law rather than ad-hoc data curation.
Researchers propose SynthAPI, a pipeline that synthesizes tool-use training environments with explicit control over tool, task and database diversity. They trained 849 models of four different sizes on 458,752 generated tasks to derive a scaling law predicting downstream agent generalization from data scale and composition. The results show that, as data budgets grow, scaling tool and database diversity beats adding more tasks, and higher data complexity yields larger gains. Agents trained under these findings outperform models trained on data from prior work across four evaluation benchmarks.
Scaling Laws for Tool-Use Agents via Controllable Environment Synthesis
We explore how to build performant large language model (LLM) tool-use agents that complete complex, multi-step tasks. Specifically, we study how data composition in synthetically generated tool-use environments affects the training of such agents. To do so, we propose the SynthAPI pipeline, which explicitly controls the diversity of generated environments along the axes of tools, tasks, and the databases on which the tools operate. By systematically varying SynthAPI's generation controls in 458,752 tasks across 849 trained models of four different sizes, we derive a scaling law that predicts downstream tool-use agent generalization based on the scale and composition of the synthesized environments. These predictable relationships reveal several lessons for scaling tool-use agent data. As data budgets grow, scaling tool and database diversity becomes more effective than increasing the number of tasks, and increasing data complexity yields larger gains. Scaling according to these findings yields agents that outperform models trained on data from prior work in large-scale tool-use agent and synthetic environment generation across four evaluation benchmarks.