Snowflake, OpenAI agree USD $200m enterprise AI deal

Snowflake has formalised a USD $200 million partnership with OpenAI, a move designed to integrate OpenAI's advanced models directly into Snowflake's suite of products and services.

This collaboration aims to provide Snowflake's global customer base, which spans all major cloud platforms, with seamless access to sophisticated artificial intelligence tools. By embedding these models natively within Snowflake Cortex AI, the company ensures that its 12,600 enterprise clients can leverage generative AI capabilities without moving their data across different environments.

The two organisations have committed to a deep level of cooperation, focusing on the co-development and deployment of customised AI solutions tailored for joint enterprise customers. Beyond product integration, the agreement encompasses extensive joint engineering efforts and a unified go-to-market strategy. This multi-cloud approach ensures that the partnership's benefits are accessible regardless of whether a client utilises Amazon Web Services, Microsoft Azure, or Google Cloud Platform, effectively streamlining the adoption of AI for large-scale organisations.

Product access

Snowflake said OpenAI models, including GPT-5.2, will be accessible within Snowflake Intelligence. Snowflake described Snowflake Intelligence as an enterprise intelligence agent that lets employees query organisational knowledge using natural language.

The company said the integration will allow customers to run AI workloads alongside proprietary data held in Snowflake. Snowflake also pointed to governance controls in Snowflake Horizon Catalog.

Snowflake said its platform includes business continuity and disaster recovery measures. It also cited a 99.99% uptime service-level agreement.

"By bringing OpenAI models to enterprise data, Snowflake enables organizations to build and deploy AI on top of their most valuable asset using the secure, governed platform they already trust," said Sridhar Ramaswamy, CEO, Snowflake.

Engineering plans

Joint engineering teams from both companies are set to collaborate on developing new features that leverage OpenAI's Apps SDK, AgentKit, and various APIs, with a primary focus on streamlining shared enterprise workflows. To facilitate this, Snowflake has introduced a suite of specialised functions within Cortex AI designed to apply OpenAI models across a diverse range of data formats.

These capabilities extend to both structured and unstructured data, encompassing text, images, and audio, allowing technical teams to explore and analyse complex datasets using standard SQL.

According to the organisations, this strategic alliance positions OpenAI as a primary model provider within the Snowflake platform. By integrating these tools so deeply into the existing infrastructure, the partnership aims to lower the barrier for businesses looking to build agentic workflows and sophisticated AI applications directly where their data resides.

This technical alignment underscores a broader effort to make high-performance generative AI a foundational element of the Snowflake data ecosystem.

"Snowflake is a trusted platform that sits at the center of how enterprises manage and activate their most critical data," said Fidji Simo, CEO of Applications at OpenAI.

"This partnership brings our advanced models directly into that environment, making it easier to deploy AI agents and apps, so businesses can close the gap between what AI is capable of and the value they can create today," said Simo.

Customer use

Snowflake highlighted Canva and Whoop as customers that plan to use OpenAI models within Snowflake. The companies framed the partnership around so-called context-aware AI applications and agents that operate over enterprise data and workflows.

Canva set out its interest in applying OpenAI models within Snowflake Cortex AI as it expands its visual AI offering.

"As we scale our visual AI offering on Canva, both OpenAI and Snowflake have played key roles in how we rapidly empower our users with new creative tools," said Helen Crossley, Head of Data Science, Canva.

"As our platform continues to scale, Snowflake has been foundational to how we manage and activate data, and we're excited to explore how leveraging OpenAI models in Snowflake Cortex AI can help us extend that foundation," said Crossley.

"The ability to bridge advanced AI models with our enterprise data allows us to move quickly and test new ideas, without compromising on security or performance," said Crossley.

Whoop said it has rolled out Snowflake Intelligence internally and has used Cortex Agents for analytics and decision-making.

"Speed and precision in decision-making are critical for us as WHOOP continues to scale," said Matt Luizzi, Senior Director of Business Analytics, WHOOP.

"Rolling out Snowflake Intelligence to our employees and developing Cortex Agents has provided a secure and governed way for WHOOP to analyze data and make decisions," said Luizzi.

"With OpenAI's models available directly within Snowflake Cortex AI, we can further enhance those agents with advanced reasoning and analysis, all while maintaining strong security and governance," said Luizzi.

"This partnership will help us continue to make AI a practical, everyday tool for the business," said Luizzi.

Existing ties

The companies noted that this agreement builds upon an established relationship, highlighting that OpenAI already utilises Snowflake's platform for experiment tracking, analytics, and testing. Internally, Snowflake has adopted ChatGPT Enterprise as a core component of its strategy to enhance workforce productivity and encourage the daily integration of artificial intelligence tools among its employees.

Looking ahead, Snowflake and OpenAI intend to broaden their collaborative product development. This expansion is expected to include new features that more deeply connect OpenAI's specialised tooling with Snowflake's robust data platform and comprehensive governance services.

By aligning these technologies, the partners aim to ensure that enterprise AI deployment remains both secure and highly integrated with existing data management frameworks.