Anthropic and Snowflake announced a deal that will allow the deployment of AI agents capable of complex, multi-step analysis inside Snowflake's governed data environments.

The combination is a win for Anthropic's efforts to get its large language models (LLMs) used in enterprises ahead of other providers like OpenAI and Google. It will allow customers in the highly regulated finance and healthcare industries – among others – to deploy agents and then pull answers from across an enterprise’s Snowflake instance with greater than 90 percent accuracy on complex text-to-SQL tasks, the companies claimed in the [1]deal [2]announcements published Wednesday. The press release said the deal was worth $200 million, but it's not clear who's paying whom nor how that number is measured. We've reached out to both companies for more info and will update this story if they respond.

It's worth noting that anything less than 100 percent accuracy means humans will still need to be involved to make sure the answers delivered are correct and actionable, as is [3]nearly always the case with AI tools.

[4]

The service will be available to more than 12,600 Snowflake customers on Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Azure.

[5]

[6]

"Enterprises have spent years building secure, trusted data environments, and now they want AI that can work within those environments without compromise," said Dario Amodei, CEO and Co-Founder of Anthropic in a canned statement. "This partnership brings Claude directly into Snowflake, where that data already lives. It's a meaningful step toward making frontier AI genuinely useful for businesses."

By bringing Claude to Snowflake, customers can get natural-language AI answers from structured and unstructured data while keeping everything within Snowflake's security and governance perimeter. This includes multimodal data analysis that uses Snowflake’s Cortex AI Functions and Claude’s Opus 4.5 to query text, images, audio, and tables, using SQL.

[7]Anthropic reduces model misbehavior by endorsing cheating

[8]Snowflake goes all out to woo PostgreSQL developers with lakehouse extensions

[9]If you want a picture of the future, imagine humans checking AI didn't make a mistake – forever

[10]Snowflake and Databricks bank PostgreSQL acquisitions to bring transactions onto their platforms

The partnership will also allow customers to use Snowflake Cortex Agents to build production-ready data agents that are powered by Claude to “retrieve and reason” across structured and unstructured data. Snowflake said these multi-agent solutions can work with and learn context from financial, operational and customer data in a way where it shows its work while it retrieves an answer.

As an example, the data-platform biz said customers in finance can use its Snowflake Intelligence with Claude to generate portfolio recommendations for clients that take into account the client's holdings, current market data and compliance rules.

[11]

“Anthropic joins a very select group of partners where we have nine-figure alignment, co-innovation at the product level, and a proven track record of executing together for customers worldwide. Together, the combined power of Snowflake and Claude is raising the bar for how enterprises deploy scalable, context-aware AI on top of their most critical business data,” Snowflake CEO Sridhar Ramaswamy said in a statement that accompanied the announcement.

In the announcement, Snowflake said its customers are already using its own AI platform, Cortex AI, which uses Claude on the backend to process trillions of tokens per month. ®

Get our [12]Tech Resources



[1] https://www.snowflake.com/en/news/press-releases/snowflake-and-anthropic-announce-200-million-partnership-to-bring-agentic-ai-to-global-enterprises/

[2] https://www.anthropic.com/news/snowflake-anthropic-expanded-partnership

[3] https://www.theregister.com/2025/07/16/if_you_want_a_picture/

[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2aTISiX_y7R55PK-AJ0YyJQAAAMU&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0

[5] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aTISiX_y7R55PK-AJ0YyJQAAAMU&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0

[6] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33aTISiX_y7R55PK-AJ0YyJQAAAMU&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0

[7] https://www.theregister.com/2025/11/24/anthropic_model_misbehavior/

[8] https://www.theregister.com/2025/11/05/snowflake_postgresql_push/

[9] https://www.theregister.com/2025/07/16/if_you_want_a_picture/

[10] https://www.theregister.com/2025/06/10/snowflake_and_databricks_bank_postgresql/

[11] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aTISiX_y7R55PK-AJ0YyJQAAAMU&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0

[12] https://whitepapers.theregister.com/