Meta launches Muse Wonder: what Shopify merchants can test

Meta’s new model is available through its API and coding tools. The Muse app upgrade is still coming; here is what Shopify teams can test.

Will Slatton2 min read

Meta launched Muse Wonder and Muse Wonder Full on 12 October, with access through its developer and coding products first. The company says Wonder will reach the Muse personal agent app soon, starting with subscribers. That distinction matters for Shopify merchants: a model release is not confirmation that every shopper’s agent has already upgraded.

Meta’s launch announcement lists Muse Wonder in the Meta Model API and Muse Code today. Wonder Full is available to Muse Code subscribers. It gives no firm date for the consumer app rollout.

What Meta says is different

Meta describes Active Memory as retrieving relevant context across conversations, and Active Focus as persisting through longer tasks, trying alternative approaches and checking in before important actions. These are vendor descriptions of model capabilities, not evidence of a measured increase in completed ecommerce orders.

💡
Availability: the API and coding release is live. Meta says the Muse app upgrade is coming soon, starting with subscribers; a general consumer rollout date has not been announced.

A Shopify founder’s early reaction

“This model is very very good.”

Shopify CEO Tobi Lütke posted that assessment on 12 October. His enthusiasm is an early personal evaluation, not a controlled benchmark or an announcement of a new Shopify integration.

What to test when access arrives

Our recommendation is to use a small, repeatable shopping task rather than redesigning a store around launch claims. Ask the agent to identify a particular variant, state its current price and availability, and explain shipping and returns. Compare each answer with the store’s own information before authorising a purchase.

Record the model and access level used, the task, and any incorrect answers. Repeat the same test after the app upgrade becomes available. That gives you a useful comparison without attributing every improvement or failure to the new model.

App teams can evaluate the API release now against a representative task set. Measure successful completion, required human corrections and total cost per task before deciding whether to switch providers.

Meta’s announcement includes researcher videos and demonstrations. These illustrate Meta’s intended capabilities; they are not independent shopping tests.

Filed underNews

More from News

See all →