Anthropic and Snowflake have expanded their partnership under a new multi-year, $200 million agreement that deepens the integration of Anthropic’s Claude models into Snowflake’s enterprise data platform and formalizes a joint go-to-market effort focused on deploying AI agents at scale.
Under the expanded arrangement, Claude models are available to Snowflake’s more than 12,600 customers through Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Azure. The integration is designed to allow enterprises to apply large language models directly to governed data environments, combining structured and unstructured data while maintaining existing security, access controls, and compliance requirements.
The partnership reflects a broader shift in enterprise AI adoption, moving from standalone model access toward tightly integrated deployments that sit alongside core data infrastructure. For large organizations, the ability to run advanced reasoning models where data already resides reduces data movement, simplifies governance, and lowers operational risk—key constraints that have slowed production deployment of generative AI in regulated environments.
Snowflake is also using Claude internally across engineering and commercial operations. Within its engineering organization, Claude Code is used to support software development workflows. On the commercial side, Snowflake has deployed a Claude-powered go-to-market assistant built on Snowflake Intelligence, enabling sales teams to query centralized data using natural language and accelerate deal execution.
At the platform level, Snowflake reports that thousands of customers already process trillions of Claude tokens per month through Snowflake Cortex AI. The next phase of the partnership centers on AI agents designed to perform multi-step analysis across enterprise data. In this model, business users pose questions in natural language, while Claude determines which datasets are required, retrieves them from across the Snowflake environment, and produces an answer. Snowflake says internal benchmarks show more than 90% accuracy on complex text-to-SQL tasks, a key requirement for enterprise analytics workloads.
For organizations in financial services, healthcare, and life sciences, the combination of advanced reasoning with a governed data layer is positioned as a way to move beyond pilots and into production deployments. Governance, auditability, and observability are embedded through Snowflake’s Horizon Catalog, which provides lineage, access controls, and responsible AI features intended to support regulatory compliance.
Dario Amodei, CEO and co-founder of Anthropic, said: “Enterprises have spent years building secure, trusted data environments, and now they want AI that can work within those environments without compromise. This partnership brings Claude directly into Snowflake, where that data already lives.”
Sridhar Ramaswamy, CEO of Snowflake, framed the deal as a deep product-level collaboration rather than a distribution agreement. He said the partnership reflects “nine-figure alignment, co-innovation at the product level, and a proven track record of executing together for customers worldwide,” with a focus on deploying scalable, context-aware AI on critical enterprise data.
From a product perspective, Claude Sonnet 4.5 now powers Snowflake Intelligence, which functions as an enterprise intelligence agent capable of answering questions across structured tables and unstructured content. Through Snowflake Cortex AI Functions, customers can also use Claude models, including Claude Opus 4.5, to query text, images, audio, and tabular data directly using SQL. Snowflake Cortex Agents extends this capability by allowing teams to build custom, production-grade AI agents that retrieve and reason over enterprise data with built-in performance and accuracy controls.
Early customer deployments highlight how these capabilities are being applied in practice. Simon Data, a composable customer data platform provider, uses Claude within Snowflake to identify patterns and relationships in customer data while operating within strict governance constraints. Intercom uses Claude through Snowflake Cortex AI to support its Fin AI Agent for customer service. According to Dave Lynch, Intercom’s vice president of engineering, the integration has increased automation rates and improved efficiency in large, complex customer environments.
The partnership also illustrates potential use cases in highly regulated domains. A wealth management firm, for example, can use Snowflake Intelligence powered by Claude to synthesize client portfolios with market data and compliance rules to generate personalized recommendations, all within Snowflake’s security and governance perimeter.
Taken together, the expanded Anthropic–Snowflake partnership underscores a growing emphasis in enterprise AI on deployment maturity: integrating advanced models with existing data platforms, enforcing governance by default, and enabling AI systems that can reason across complex datasets rather than simply generate text.