Shopify has announced a multi-year partnership with Liquid AI, licensing their Liquid Foundation Models (LFMs) to power core commerce experiences,especially for search and personalized product recommendations.
The first production deployment: a sub-20ms text model that dramatically enhances the speed and relevance of Shopify store search.
On top of that, Shopify and Liquid co-developed a generative recommender system with a novel HSTU architecture. In tests, this system outperformed Shopify’s prior stack boosting recommendation quality and potentially improving conversion.
Why it matters:
- Low latency: LFMs are designed for real-time inference, making the search feel instant and fluid.
- Efficiency: These models use fewer parameters than many open models (like Qwen3 or Llama 3) while maintaining or improving quality.
- Bigger vision: Shopify and Liquid are eyeing future multimodal use cases not just text for customer profiles, recommendation agents, and product classification.
This partnership signals Shopify’s deepening commitment to AI at the infrastructure level , not just for merchant tools, but embedded in the platform’s core. For growth marketers, community leaders, and founders, this means smarter search, better recommendations, and potentially more personalized shopper journeys.