Ali Ghodsi, founder and CEO of Databricks, speaks at the company’s Data and AI conference in San Francisco on June 16, 2026.
Grant Terzakis Media | Databricks
Databricks has a unique role in the AI boom. Revenues continue to rise as businesses flock to the company’s data analytics tools. But as clients rely on multiple AI agents to clean data and ask questions, Databricks endures higher costs, resulting in lower margins.
“It’s a consumption-based business model, agent AI is coming,” Databricks CEO Ali Ghodsi told CNBC in an interview Tuesday at the company’s Data and AI conference in San Francisco. “Agents generate a lot of inquiries. We have all these agents, agents and a platform of agents that we have that generate revenue, so it increases the use of everything around.”
Databricks told analysts at the conference that annual revenue jumped more than 80% from a year ago and now stands at $6.9 billion, up from $5.4 billion in the fourth fiscal quarter.
With a standalone market value of $134 billion, Databricks is worth more than its competitors Snowflakewhich went public in 2020 and now has a market cap of around $83 billion. Snowflake’s annual revenue is estimated at $5.6 billion, based on its latest quarterly results published last month.
Databricks continues to sit on the sidelines of the public market as its more respected peers pursue IPOs. SpaceX It hit its biggest hit in history last week, reaching $2 trillion on its first day of trading on Friday. And AI model developers OpenAI and Anthropic have applied for a private offering.
Although Databricks is often associated with AI modeling companies, it has a very different role in the market. Databricks’ Genie can answer questions from business users about business data, and its Agent Bricks allow developers to build custom AI applications. Since those products are so popular, Databricks has to spend more on the sub-models.
Ghodsi declined to give Databricks’ current gross, but said it will decline. His company now generates $1.7 billion in annual revenue from AI products, up from $1.4 billion in February.
One of the biggest trends in AI today is that companies are cutting back on the use of tokens. Databricks’s Unity AI Gateway can inform people as they get closer to spending their AI budget.
Companies have moved away from tokenmaxxing, or encouraging employees to use as many tokens as possible, and are now “value-maxxing” to improve efficiency while still harnessing the power of AI, Ghodsi said.
Big companies “want to fully utilize the best, most intelligent models,” he said, highlighting Anthropic’s Mythos. “They are interested in that, but not in everything, right? And because of the common tasks, they really want to reduce costs and use simple open source models.”
Chinese models are very popular among Databricks customers, Ghodsi said.
“Customers really want choice,” he said.
Databricks is looking to grow by selling tools tailored to specific industries. The company announced its entry into the cybersecurity market in March with the launch of Lakewatch software. On Tuesday, Databricks said it would buy Panther, a security startup valued at $1.4 billion in 2021, and unveiled its CustomerLake marketing data management software.
WATCH: Databricks CEO Ali Ghodsi: AI does not have an intelligence problem. It has a context problem
