Snowflake has announced dynamic model routing within Cortex AI Gateway and Snowflake's flagship AI products, alongside expanded access to leading open models. The innovations help enterprises reduce unnecessary AI spend and improve intelligence efficiency, measuring how effectively companies turn compute, models, data, and context into business impact.
Snowflake introduces dynamic model routing in Cortex AI Gateway.
Automatically selects optimal model based on quality, speed, cost, and preferences.
Adds DeepSeek-V4-Flash and GLM-5.3 to Snowflake Cortex AI.
Internal testing shows up to 3x greater token efficiency with model routing.
Provides spend controls with quotas, limits, and usage tracking.
DeepSeek-V4-Flash scored 74.4% on data engineering tasks in testing.
With the addition of dynamic model routing, Cortex AI Gateway can automatically select the specific model with the optimal balance of quality and cost for the task at hand. Dynamic model routing is also integrated across Snowflake's flagship AI products, including Snowflake CoCo and Snowflake CoWork, and is available to third-party AI agents using Cortex AI Gateway. This new capability directs lower-complexity or repetitive tasks to more efficient models, while work that requires deeper reasoning is routed to frontier models. Snowflake's internal testing indicates that using a mix of open and proprietary models can deliver comparable quality while materially improving token efficiency, with agents using dynamic model routing building a dbt pipeline with up to 3x greater token efficiency than a frontier-model-only path.
Snowflake will be expanding customers' access to leading open models, including DeepSeek-V4-Flash 0731 and GLM-5.3, through Snowflake Cortex AI. Together with models from providers including Anthropic, OpenAI, Google, SpaceXAI, Meta, and Mistral, this portfolio gives customers more freedom to choose the right combination of performance and cost. Snowflake's AI Research Team evaluated DeepSeek-V4-Flash on enterprise-focused tasks, with recent testing showing it outperformed leading proprietary models, scoring 74.4 percent on data engineering tasks.
"Enterprises are becoming much more rigorous about the economics of AI. The question is no longer how much AI they are using, but whether that AI is translating into meaningful business value," said Sridhar Ramaswamy, CEO, Snowflake. "Achieving intelligence efficiency requires the flexibility to use the best model for each task as the landscape evolves. Snowflake's role is to absorb that complexity so customers can focus on outcomes while we optimize model choice underneath."
About Snowflake
Snowflake is the platform for the AI era, making it easy for enterprises to innovate faster and get more value from data. More than 13,900 customers around the globe, including hundreds of the world's largest companies, use Snowflake's AI Data Cloud to build, use, and share data, applications, and AI, in a secure and governed fashion.
