Snowflake, Anthropic boost partnership with $200M commitment

Better together

Over the past three years, many enterprises have substantially increased their investments in developing AI applications.

The real value of this relationship is expressed, not by the dollar amount, but by the commitment of precious resources to each other. There are only a limited number of LLM providers that get to operate directly within the Snowflake platform, and Anthropic is one of them.

OpenAI's November 2022 launch of ChatGPT marked a significant improvement in generative AI (GenAI) technology. Given GenAI's potential to make employees better informed and more efficient, enterprises quickly responded. In turn, data management vendors responded to their customers' increased interest in AI by building environments within their platforms that simplify combining proprietary data with GenAI models.

Tech giants AWS, Google Cloud and Microsoft created GenAI development environments. So did Databricks, Snowflake, Teradata and a host of others. As part of those environments, vendors made models from multiple AI developers available.

Snowflake's initial partnership with Anthropic made Claude models available in Cortex AI alongside models from other AI developers. The expanded partnership, marked by the $200-million commitment, adds technological collaboration.

Using a portion of the committed capital, Snowflake aims to build agents powered by Claude's reasoning capabilities that enable users to perform multi-step data analysis. In addition, with Anthropic's models natively available in Cortex AI, Snowflake users can build their own agents powered by Claude using the Snowflake Cortex Agents development framework.

Like Menninger, Mike Leone, an analyst at Omdia, a division of Informa TechTarget, noted that the financial commitment is less significant than the joining of forces beyond merely making Claude models available in Snowflake.

"The real story is that Snowflake is acknowledging that … it must tightly couple with a best-in-class model like Claude to bridge the reliability gap in agentic AI workflows," he said. "Snowflake is ensuring its platform remains the primary orchestration layer for governed AI and ensures the heavy lifting of agentic AI happens directly within its platform."

The financial commitment, meanwhile, serves as insurance to Snowflake customers that they will have dedicated AI inference capabilities from Anthropic, Leone continued.

"The money likely funds dedicated engineering resources to optimize Claude specifically for Snowflake's infrastructure," he said. "This could ensure Anthropic prioritizes enterprise-grade capabilities that are relevant to Snowflake, like SQL generation or schema reasoning, rather than just general-purpose conversational improvements."

While Snowflake has partnerships with numerous model providers, the expanded partnership with Anthropic was driven by the customer response Snowflake is seeing to Anthropic's models within Cortex AI, according to Dwarak Rajagopal, Snowflake's vice president of AI Research and Engineering.

"Anthropic's ability to deliver transparent, context-aware analysis at massive scale has made it a leader in enterprise AI," he said. "By deepening our partnership with Anthropic, we're making it even easier for our customers to tap into the power of Claude and deploy enterprise intelligence agents on Snowflake."

A particular focus of the $200 million partnership will be to provide capabilities that enable enterprises in highly regulated industries such as financial services and healthcare to benefit from agentic AI, Rajagopal continued.

"The investment will serve as the fuel for our multi-year collaboration efforts … particularly in regulated industries in the coming years," he said.

Anthropic's Claude models have gained a reputation as some of the highest-performing LLMs in benchmark testing. As a result, Anthropic has attracted significant funding over the past year, raising $3.5 billion in March and $13 billion in September.

Given the performance of Anthropic's models and the continued development that massive funding rounds will enable, Anthropic is an attractive partner for data management vendors such as Snowflake and Databricks, according to Menninger.

"Anthropic has built a reputation as a preferred LLM provider for code generation, and they have also focused energy on agentic AI processing, for example, with the introduction of the Model Context Protocol," he said. "Much of database processing involves SQL, which can be generated by LLMs, so it's not surprising to see them succeeding with data management vendors."

Leone, meanwhile, noted that beyond technological prowess, Anthropic is becoming a popular partner for data managers because of its single-mindedness.

"Anthropic is becoming a go-to partner because simply put, it stays in its lane," he said. "Unlike several other big model providers that are also trying to sell you on their cloud ecosystem or their cloud data platform, Anthropic … just wants to provide the underlying model. That neutrality means a vendor like Snowflake can plug in a world-class model without worrying that their partner is competing with them."

While the $200 billion partnership has specific goals, including improving agentic AI capabilities and marketing initiatives, it more broadly represents Snowflake's continued commitment to AI development.

The vendor was slower to react to ChatGPT's launch than Databricks and other competitors. As a result, when Sridhar Ramaswamy was named Snowflake's new CEO in February 2024, the vendor had to catch up to its peers or risk becoming outdated.

Throughout 2024 and 2025, Snowflake has aggressively added and improved its AI development capabilities. However, it still has catching up to do before it matches what Databricks, AWS, Google Cloud and Microsoft provide, according to Menninger, who helps produce ISG Research's annual Buyers Guide for AI Platforms.

"ISG Research rated Snowflake exemplary … but still lagging behind Databricks and the hyperscalers," he said. "The new battlefield for all these competitors is AI agents, so anything these providers can do to make agent authoring, deployment and evaluation easier will help on the competitive front."

Next steps

As Snowflake looks ahead to 2026, the vendor's primary focus is on further enabling customers to develop and deploy AI tools in the same environment their data resides, according to Rajagopal.

"We'll continue doubling down on agentic AI so every organization can retrieve, reason over and act on all their structured and unstructured data securely inside Snowflake," he said. "You'll also see us push hard on AI interoperability -- embracing open standards like Model Context Protocol so customers can use any model and connect agents across their ecosystem."

Menninger, meanwhile, suggested that Snowflake build more agents into its platform to help users discover and prepare data for AI development.

"The battle is just beginning, but it absolutely includes making data available and usable for AI," he said. "More than half the participants in our research cite this as their biggest data and AI challenge. More built-in agents to handle data engineering tasks would be a great way to deliver more value to customers."

Eric Avidon is a senior news writer for Informa TechTarget and a journalist with more than 25 years of experience. He covers analytics and data management.