Budget-Aware Tool Use Enables Effective Agent Scaling
Quick summary
arXiv:2511.17006v2 Announce Type: replace Abstract: Scaling test-time computation has been extended from language model reasoning to tool-augmented agents, where scaling involves not only thinking in tokens but also acting via tool calls that directly constrain environmental interaction. However, we found that simply increasing the tool-call budget fails to improve performance, as agents lack "budget awareness" and quickly hit a performance ceiling. We study how to scale such agents effectively under explicit tool-call budgets, focusing on web search agents. We first introduce the Budget Track
Key takeaways
- arXiv:2511.17006v2 Announce Type: replace Abstract: Scaling test-time computation has been extended from language model reasoning to tool-augmented agents, where scaling involves not only thinking in tokens but also acting via tool calls that directly constrain environmental interaction.
- However, we found that simply increasing the tool-call budget fails to improve performance, as agents lack "budget awareness" and quickly hit a performance ceiling.
- We study how to scale such agents effectively under explicit tool-call budgets, focusing on web search agents.
Why it matters
This development shows AI moving deeper into everyday software. Productivity potential should be weighed against price, data permissions, exportability and the preservation of human control.
Member comments