Snowflake adds dynamic model routing to Cortex AI Gateway

Snowflake has launched dynamic model routing within its Cortex AI Gateway, alongside expanded access to open-source AI models, as the company looks to help enterprises reduce unnecessary AI spend and improve what it calls "intelligence efficiency".

The new capability allows Cortex AI Gateway to automatically select the model that offers the optimal balance of quality and cost for a given task. Lower-complexity or repetitive tasks are directed to more efficient models, while work requiring deeper reasoning is routed to frontier models.

Snowflake said the feature removes the need for development teams to manage model selection for each individual request.

Dynamic model routing is integrated across Snowflake's flagship AI products, including Snowflake CoCo and Snowflake CoWork, and is also available to third-party AI agents using Cortex AI Gateway.

Snowflake said administrators retain control over which models and providers are available to their users - a consideration it flagged as important for regulated industries and global organisations navigating regional model availability requirements.

Snowflake's internal testing found that agents using dynamic model routing built a dbt pipeline with up to three times greater token efficiency than a frontier-model-only approach, while maintaining comparable quality. In a separate test, the company said engineering teams completed the same number of pull requests with 25% greater token efficiency.

Alongside the routing update, Snowflake said it would expand access to open models through Snowflake Cortex AI, adding DeepSeek-V4-Flash 0731 and GLM-5.3 to its model library.

The portfolio already includes models from Anthropic, OpenAI, Google, SpaceXAI and Mistral. Snowflake's AI Research Team evaluated DeepSeek-V4-Flash on enterprise-focused tasks, with recent testing showing it scored 74.4% on data engineering tasks, which the company said outperformed leading proprietary models tested in the same evaluation.

GLM-5.2 scored 62.8% while using fewer tokens than any other model tested, according to Snowflake.

Because both new models are open source and can be self-hosted, Snowflake said organisations can keep all data within their own environment. Open and proprietary models are subject to the same access controls, security policies and governance framework within the platform.

Cortex AI Gateway also gives administrators visibility into token usage and costs, and allows organisations to set spending limits across AI apps and agents. Snowflake CoCo extends those controls through role-based access and tagging, enabling per-user quotas, cost-centre attribution and consumption notifications.

"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, chief executive at 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 optimise model choice underneath."

Snowflake announced Cortex AI Gateway in July 2026 as a unified foundation for governing agent connections, routing requests and optimising AI consumption.