On 2026-10-22, NVIDIA published an account of European retailers using its open models and physical AI tools, led by a commerce platform that it says cut inference costs by 90% (NVIDIA). Shopify post-trained NVIDIA Nemotron 3.5 Lightning for the GraphQL subagent in Sidekick, its AI assistant for merchants (NVIDIA). On Shopify’s internal benchmark, the post-trained model surpasses a leading frontier model in accuracy while cutting serving costs by 90% and delivering responses 7x faster, according to NVIDIA (NVIDIA).
Key Facts
| Item | Detail | Source |
|---|---|---|
| Shopify | 90% lower serving costs and 7x faster responses on its internal benchmark | NVIDIA |
| Zooplus | 86% search accuracy in a production ready prototype on AWS | NVIDIA |
| Vestiaire Collective | About 40,000 images a day standardized across 70 countries | NVIDIA |
| CATCHES with AMIRI | Conversion up 55%, add-to-cart up 80%, gross sales up 56% | NVIDIA |
| Mirakl | More than 70 million products onboarded | NVIDIA |
Timeline
- 2026-10-22: NVIDIA published its account of European retail deployments of open models and physical AI (NVIDIA).
Details (as Reported)
Shopify’s engineers use the NVIDIA Megatron Bridge library for faster fine-tuning and NVIDIA NeMo RL for reinforcement learning (NVIDIA). NVIDIA reported that its CCluster vLLM inference optimization software helped Shopify reach 4.5x faster performance and up to 12x lower inference cost across five additional use cases (NVIDIA). Those use cases are catalog enrichment, merchant assistant, e-commerce shopping simulation, product taxonomy classification, and merchant and buyer recommendations (NVIDIA).
Zooplus worked with Infosys and NVIDIA technologies to rebuild its product discovery stack, using the NVIDIA Nemotron Retriever library for semantic search (NVIDIA). Photoroom’s Image Editing API now standardizes every primary image for Vestiaire Collective at the point of listing, and listings it processed sold at a 5% higher rate of liquidity, with 3% to 7% more unique views per product page (NVIDIA).
Omi launched ProductDrop, which anchors every generated image to an approved digital twin, and L’Oréal is using the approach within its CreAItech program (NVIDIA). L’Oréal is also integrating NVIDIA ALCHEMI into its research, moving from months of experimentation to hours of computational screening (NVIDIA). Orbital Industries’ OrbMol-v2 is five times faster than comparable models and delivers 21% lower error than the next-best model of its kind, per NVIDIA (NVIDIA).
Otto Group advanced its autonomous warehouse under human supervision initiative to its largest fulfillment center in Haldensleben, Germany, built with Reply on NVIDIA Omniverse libraries and running on Google Cloud (NVIDIA). Logistics teams can design, test and optimize warehouse operations entirely in simulation before any physical implementation (NVIDIA).