OpenAI and Databricks have signed a $100 million multiyear deal to deliver OpenAI’s flagship GPT models to enterprise clients, streamlining the creation of custom AI agents.

Under the agreement, OpenAI models will be built into Databricks’ cloud platform, being available to its 20,000-plus customers within the Databricks Data Intelligence Platform and through Databricks’ AI product, Agent Bricks.

Using their own data stored on the Databricks platform, enterprises will be able to tap into OpenAI’s latest models, including GPT-5, to build, evaluate and scale custom AI apps and agents.

The result, according to Databricks, will be production-ready AI tools, with OpenAI’s models bringing the high-capacity processing, and Agent Bricks further tuning and optimising to deliver high-quality, domain-specific outputs.

“We’re seeing overwhelming demand from enterprise customers looking to build AI apps and agents on their data, tailored to their unique business needs,” said Ali Ghodsi, co-founder and CEO of Databricks. 

“This partnership makes it easier for enterprises to securely leverage their data and OpenAI models at scale with best-in-class governance and performance.”    

While the specific financial aspects of the partnership between the two Silicon Valley giants remain under wraps, Brad Lightfoot, OpenAI’s chief operating officer, told the WSJ  that the firms expected to “far eclipse” $100 million in revenue. 

However, the terms mean Databricks will need to pay out regardless of whether customers and profits follow over the undefined timeline of the deal.   

Recommended reading

• OpenAI to Launch AI-Powered Jobs Platform in 2026
• Is the UK Falling Behind in the Race for AI Talent?
• PwC: Firms Now Pay a 56% Premium for AI Talent

Databricks said the deal kicks off a “deep engineering collaboration” with OpenAI, adding another notch to its belt following the company’s March agreement with Anthropic to deliver Claude models to over 10,000 firms for AI agent development.

Databricks has been on something of a spree this year, having inked similar deals with Palantir, Microsoft, and, just days ago, the London Stock Exchange Group in a partnership that will allow financial teams to build governed AI agents on their own data.

The company seems to reckon it has found a recipe for success in this model, and that might be true given that this month Databricks closed its Series K funding, raising $1 billion of capital at a valuation of over $100 billion, perhaps built on expectations that it will soon make more than a billion from its AI offerings.