New embedding model released with improved capabilities and lower pricing

AI Impact Summary

A new embedding model has been released that is described as more capable, cost-effective, and simpler to use. This pricing-focused update creates a clear opportunity to reduce costs for embedding workloads, but teams should benchmark the new model against the current one to understand any changes in quality or throughput and adjust downstream pipelines accordingly. Plan to update integrations to use the new model and revalidate vector store indexing and downstream consumers to realize the full cost and usability benefits.