Shopify post-trains open-source model as GraphQL task expert for Sidekick
Overview
Shopify's Sidekick assistant uses a frontier model as orchestrator and a post-trained open-source model as its GraphQL sub-agent, according to a post by X user meng shao (@shao__meng) reporting Shopify's description.
The post says the open model's judge score rose from 37.7 to 74.5, against 73 for the frontier model on GraphQL tasks. Median response time is 0.87 seconds, versus 6.50 seconds through the frontier API, and self-hosting cuts cost about 90%. These figures come from the post, and the report does not include Shopify's own published benchmark.
Written by AI from the articles below · updated Oct 21, 9:04 AM ET
Check the sources:
Article timeline
The articles in this story. Times are ET.
meng shao@shao__mengXPickShopify trains open model into GraphQL task expert, cutting cost about 90%AIShopify's Sidekick assistant pairs a frontier model as orchestrator with a post-trained open-source model for GraphQL sub-agent tasks. The post reports the open model's judge score rose from 37.7 to 74.5, against 73 for the frontier model. Median response time is 0.87 seconds versus 6.50 seconds through the frontier API, and self-hosting cuts cost about 90%.

Heat Trend
Not enough continuous observations to show a trend yet.