Snowflake brings Google Gemini 3 AI models to Cortex

Snowflake has expanded its collaboration with Google Cloud and will offer Google's latest Gemini 3 large language models natively inside Snowflake Cortex AI, in a move aimed at closer product integration and joint sales activity between the two companies.

The arrangement links Snowflake's AI Data Cloud with Google's generative AI stack and infrastructure. It allows joint customers to run Gemini 3 against data stored and governed within Snowflake, rather than exporting information to external AI services.

Snowflake said customers can develop and deploy generative AI applications directly in Snowflake Cortex AI using Gemini models. They can also build so-called intelligent data agents that sit on top of enterprise data without copying that data across platforms.

The company positioned the move as a way for enterprises to apply generative AI to existing analytics and data management workloads while keeping information in a single controlled environment. It said this structure supports security, compliance and performance requirements.

Christian Kleinerman, EVP of Product at Snowflake, said the collaboration builds on Google's role in large-scale cloud and AI infrastructure.

"Google has long set the standard for innovation and large-scale infrastructure, remaining at the forefront of the generative AI revolution," said Christian Kleinerman, EVP of Product, Snowflake. "By combining Google Cloud's industry leadership with Snowflake's ability to bring AI directly to enterprise data, we're empowering customers to move faster, innovate more freely, and redefine what's possible through data and AI."

The companies are also extending their go-to-market alignment. Joint customer engagements will increase, along with new co-selling arrangements and the option for customers to transact through Google Cloud Marketplace.

Snowflake plans to launch its service on Google Cloud in additional regions. It is now live on Google Cloud in the Kingdom of Saudi Arabia and is preparing an expansion to Google Cloud in Melbourne, Australia, which is scheduled for early 2026.

The partners are focusing on running Snowflake workloads more broadly on Google's infrastructure. Snowflake Gen2 Warehouses are now available in production on Google Cloud C4A virtual machines based on the Axion processor line. Snowflake describes this as a significant infrastructure upgrade that changes the price‑performance profile of its warehousing service on Google Cloud.

Michael Gerstenhaber, Google Vice President, Agents, API & Vertex AI, said the technical integration brings Google's models closer to customer data held in Snowflake.

"Through this partnership, Google Cloud and Snowflake are empowering enterprises to innovate faster," said Michael Gerstenhaber, Google Vice President, Agents, API & Vertex AI, Google Cloud. "By bringing Gemini models natively into Snowflake, customers can use the world's most advanced generative AI capabilities on their governed data, unlocking new insights and automation opportunities across industries."

The two companies said enterprises across sectors such as financial services, healthcare, manufacturing, retail and supply chain, technology and analytics are already using joint Snowflake and Google Cloud offerings. These customers are modernising data platforms and introducing AI into business processes.

Snowflake cited BlackLine and Fivetran as early adopters using Snowflake's AI Data Cloud on Google Cloud in production. These firms are applying Snowflake for real-time analytics, AI model deployment and data governance across their operations.

BlackLine is developing AI tools aimed at finance functions that sit on top of a combined Snowflake and Google Cloud stack. It is embedding Gemini models inside its own platform that runs on Snowflake.

"BlackLine is pioneering agentic AI for the Office of the CFO, moving beyond simple automation to create AI that actively partners with finance & accounting professionals", said Jeremy Ung, BlackLine Chief Technology Officer. "By embedding Google's powerful Gemini models into our unified data platform on Snowflake, we are empowering our platform to tackle the most complex financial processes with greater speed and intelligence than ever before."

Data integration provider Fivetran is also linking its services with the joint Snowflake and Google Cloud environment. The company focuses on creating a consolidated and governed data layer that AI services can query.

"Enterprises are finally able to do far more with their data than simply store it or run basic queries," said George Fraser, CEO of Fivetran. "Fivetran delivers the connected and governed data foundation that allows Gemini in Snowflake Cortex AI to actually think and reason with customer data. What used to take weeks of custom development now happens in days, simply by asking questions and getting answers from the data that already runs their business."

Snowflake said more than 12,000 customers now use its platform globally, including several hundred large enterprises. It expects further adoption of generative AI services on Google Cloud infrastructure as the expanded regional coverage and deeper product integrations roll out over the next year.