04:00
2026-10-12
arxiv.org
ai-agents
SWE-Duel: Towards a Self-Scaling Arena for Coding Agents
SWE-Duel, a self-scaling coding-agent arena introduced in arXiv paper 2610.12569v1, ran a 1,500-turn tournament among 11 model-harness participants on 15 repositories in five languages, with GPT-5.5 o…