88Z AI, 4 September — Alibaba and Taotian Group jointly released the E-Commerce Bench, an evaluation benchmark for e-commerce operations. It requires large models to simulate running an online store for one year using 100,000 yuan in capital based on real market data, covering the entire process including product selection, pricing, negotiation, inventory, and promotions. Tests showed that GPT-5.6 Sol led with a 14-fold asset increase. Models varied significantly in anti-fraud and cash flow management, with only Qwen
At a glance
- E-Commerce Bench requires models to simulate running an online store for one year with 100,000 yuan in real data
- GPT-5.6 Sol led with a 14-fold asset increase, while only Qwen3.8-Max-Preview could lower prices in repeat purchases
The update
Alibaba, in collaboration with Taotian Group, has officially open-sourced the E-Commerce Bench, designed to evaluate the practical capabilities of large models in complex business scenarios. This benchmark establishes a highly realistic testing environment, requiring models to operate an online store for 365 days using 100,000 yuan in start-up capital against real market data. The assessment comprehensively covers the entire e-commerce chain, including product selection strategies, dynamic pricing, supplier negotiations, inventory optimization, and promotional activities. In evaluations of 18 mainstream models, GPT-5.6 Sol topped the list with a 14-fold increase in assets. Significant differences were observed among models in key dimensions such as fraud detection and cash flow management. Notably, Qwen3.8-Max-Preview was the only model capable of achieving gradual price reductions during customer repeat purchases, demonstrating unique advantages in long-term gaming and cost control.
Sources & editorial note
This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.
- 千问大模型Source published
Source dates above refer to the original announcements; the date under the headline is this article’s publication on 88Z AI.