Snowflake announced two capabilities in Snowflake Cortex AI: dynamic model routing through Cortex AI Gateway, which selects the lowest-cost model that can complete each step of an agent task, and an expanded open-model portfolio including DeepSeek-V4-Flash (now in private preview) and GLM-5.3 (coming soon)1. Evaluated on ADE-bench using Snowflake CoCo, DeepSeek-V4-Flash scored 74.4%, outperforming the leading proprietary model Snowflake tested. In one internal evaluation, dynamic model routing completed a dbt pipeline workload with up to three times greater token efficiency than a frontier-model-only approach while delivering comparable quality.
Snowflake Adds Dynamic Model Routing and Open Models to Cortex AI
Snowflake introduced dynamic model routing and added DeepSeek-V4-Flash and GLM-5.3 to Cortex AI, with routing delivering up to 3x token efficiency gains.