Snowflake just dropped $200 million on a multi-year deal with OpenAI, and it's not even the company's first massive AI bet this quarter. The cloud data platform signed an identical $200 million partnership with Anthropic just two months ago, signaling a seismic shift in how enterprises are approaching AI infrastructure. Instead of betting everything on a single AI provider, companies are hedging their bets across multiple frontier models - a strategy that could reshape the competitive landscape and prevent the winner-take-all outcome many predicted.
Snowflake just made its second $200 million AI bet in as many months, and the pattern emerging tells us everything about where enterprise AI is headed. The cloud data giant announced Monday it's partnering with OpenAI in a deal that looks remarkably similar to the one it struck with Anthropic back in December. Same price tag, same multi-year commitment, nearly identical press release language about empowering customers with cutting-edge AI.
The deal gives Snowflake's 12,600 enterprise customers access to OpenAI models across Amazon Web Services, Microsoft Azure, and Google Cloud Platform. Snowflake employees also get ChatGPT Enterprise licenses, while both companies commit to building new AI agents and products together. "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," Snowflake CEO Sridhar Ramaswamy said in the announcement.
Here's where it gets interesting. When Ramaswamy announced the Anthropic partnership in December, he used nearly word-for-word identical language about giving customers access to powerful models on their data. That's not an accident or lazy PR - it's a deliberate strategy that Snowflake is broadcasting loud and clear.
"Our partnership with OpenAI is a multi-year commercial commitment focused on reliability, performance, and real customer usage. At the same time, we remain intentionally model-agnostic," Baris Gultekin, vice president of AI at Snowflake, told TechCrunch. "Enterprises need choice, and we do not believe in locking customers into a single provider. OpenAI is an important partner, and it is one of several frontier model providers available on Snowflake today, alongside Anthropic, Google, Meta, and others."
Snowflake isn't alone in this approach. ServiceNow, the workflow automation platform, announced multi-year deals with both OpenAI and Anthropic in January. ServiceNow president, COO and CPO Amit Zavery explained the dual partnership as giving customers and employees the ability to choose which model works best for specific tasks - essentially the same rationale Snowflake is using.
This multi-vendor trend runs counter to the conventional wisdom that's dominated AI discourse for the past two years. The narrative has been that network effects and scale advantages would create a winner-take-all market, with one or two AI providers capturing the lion's share of enterprise spending. But what's actually happening on the ground looks more nuanced and messy.
Trying to figure out who's actually winning the enterprise AI race reveals just how murky the competitive landscape remains. A Menlo Ventures survey from late 2025 showed its portfolio company Anthropic holding a commanding market lead. Meanwhile, an Andreessen Horowitz report from last week naturally found its portfolio company OpenAI leading the pack. When venture firms with skin in the game can't agree on the basics, you know the market is still wide open.
These conflicting data points actually strengthen the case for why enterprises are spreading their bets. Each AI lab's large language models come with distinct strengths and weaknesses - OpenAI's GPT models excel at certain reasoning tasks, while Anthropic's Claude handles others better, and Google's Gemini and Meta's Llama models bring their own advantages. Rather than guess which provider will dominate in three years, companies like Snowflake are building infrastructure that lets customers access all of them.
The economics make sense too. A $200 million commitment sounds massive, but for a company like Snowflake with billions in annual revenue, it's a manageable hedge against technological uncertainty. The alternative - picking one provider and getting locked in - carries far more risk if that model falls behind or if pricing becomes unfavorable.
This strategy mirrors what happened in the cloud infrastructure wars of the 2010s. Many predicted Amazon Web Services would steamroll the competition, but enterprises ultimately adopted multi-cloud strategies across AWS, Azure, and Google Cloud. The same pattern seems to be playing out with AI models, except the timeline is compressed and the stakes potentially higher.
What makes this trend particularly significant is that employees at these enterprises are already using their preferred models regardless of official contracts. IT departments are playing catch-up, trying to provide governed access to the tools workers are using anyway. By partnering with multiple providers, companies like Snowflake can meet employees where they are while maintaining security and compliance standards.
There's another angle here that the announcements don't emphasize but that's clearly part of the calculation. These partnerships give Snowflake and ServiceNow early access to the latest model releases and closer collaboration on product development. They're not just licensing technology - they're getting a seat at the table as OpenAI and Anthropic build the next generation of enterprise AI capabilities.
The enterprise AI market is shaping up to look less like a winner-take-all battle and more like the multi-cloud infrastructure landscape that emerged over the past decade. Companies are voting with their checkbooks for optionality over allegiance, spreading hundreds of millions across multiple AI providers rather than betting everything on a single horse. This could be great news for competition and innovation, forcing AI labs to keep improving rather than coasting on early leads. But it also means the decisive moment many expected - when one AI provider clearly pulls ahead - might never arrive. Instead, we're heading toward a world where enterprises juggle relationships with several AI companies simultaneously, choosing models task by task rather than making blanket commitments. For Snowflake, ServiceNow, and others adopting this strategy, it's a hedge against an uncertain future. For the AI labs themselves, it means the race is far from over.
