Alibaba.com today announced results from a 107-task benchmark evaluation showing that Accio, its AI agent platform for commerce, completed a range of e-commerce tasks at more than 50% lower estimated cost than OpenAI’s Codex and Anthropic’s Claude Code, with comparable completion quality.
This makes Accio the most cost-competitive e-commerce AI tool available to small and medium-sized businesses. The company also introduced an expanded Accio that brings research, sourcing, and daily e-commerce operations into one AI workspace for small businesses.
The announcements were made at CoCreate, Alibaba.com’s annual event for entrepreneurs and small businesses.
“For small businesses, unaffordable AI is useless,” said Kuo Zhang, President of Alibaba.com. “Our goal is not simply to make AI more powerful. It is to make commerce AI practical and affordable for even a one-person company.”
Half the cost, across 107 commerce tasks
In the 107-task benchmark evaluation, the estimated total cost of completing the full task set was $3.69 with Accio, compared with $9.27 for OpenAI’s Codex and $9.51 for Anthropic’s Claude Code — more than 50% lower in both cases.
Accio’s efficiency comes from optimising the full system around real commerce work:
- Commerce-specific intelligence. Accio uses commerce-specific data and workflows to post-train lightweight models for clearly defined tasks. This allows smaller models to handle more routine work efficiently, while more capable models remain available when deeper reasoning is required.
- The right resources for each task. Accio does not simply choose the cheapest model or default to the most powerful one. It breaks complex requests into manageable steps and balances quality, speed, cost and data requirements for each step. In economic terms, the system seeks to move each workflow closer to the Pareto frontier — where improving one dimension would require a trade-off in another. For a small business, that means getting the performance the job requires without paying for computing resources it does not need.
- Less unnecessary computation. Cache reuse, context compression and coordinated agent execution reduce repeated processing, unnecessary tool calls and redundant token consumption.
A single workspace for global e-commerce
Alibaba.com also announced that Accio has evolved into a unified workspace for global e-commerce operations. Small businesses can use Accio to research markets, identify product opportunities, develop products, evaluate suppliers and manage daily operations. Through connections with Amazon, Shopify, eBay, TikTok Shop and Walmart, sellers can access supported storefront workflows from one place.
Commerce agent bench, now open source
Alibaba.com has open-sourced Commerce Agent Bench on GitHub. Rather than relying on synthetic exercises, it draws on real merchant activity — 10 million active SMB users, 1.6 million conversations and 200,000 execution traces — distilled into 107 end-to-end commerce tasks across seven categories and four levels of autonomy. Tasks include reviewing hundreds of unstructured emails, spotting payment fraud, calculating landed costs and booking multi-carrier shipping routes.
No single model led across the board, reinforcing the case for task-level routing — the idea that different tasks are best handled by different AI models rather than relying on one general-purpose model for everything.
