Dynamic model routing expands in Cortex AI Gateway
Snowflake announced dynamic model routing in Cortex AI Gateway to cut enterprise AI inference costs through automatic model selection by workload.
The capability expands across Snowflake CoCo, Snowflake CoWork, and third-party agents; it routes simpler tasks to cheaper models and complex work to frontier models.
Internal tests showed up to 3x token efficiency on a dbt pipeline and 25% higher token efficiency for pull-request throughput.
Snowflake also disclosed expanded access to open models via Cortex AI, including DeepSeek-V4-Flash 0731 and GLM-5.3.
SNOW•
