E-Commerce Bench

"Evaluating LLM Agents on Long-Horizon Autonomous Business Operations." A Qwen Team benchmark, built with Taobao and Tmall and HKUST, in which an agent is handed ¥100,000 and up to four stores and must run them for 365 simulated days inside a deterministic market with multi-round supplier negotiation and dynamic events such as promotions, disasters, and supply shocks; the primary metric is the asset multiplier over five episodes. Eighteen models sit on the launch leaderboard. Posted August 31, 2026 with an Apache 2.0 repository (85 stars). Filed late.