Databricks Closes $5B at $190B Valuation — $25B Raised While Still Private

Enterprises are wiring AI agents into real workflows, and the money chasing the data infrastructure underneath is unlike anything the sector has seen. What stands out is where that money lives: almost none of it is in the public markets. A growing set of companies keeps stacking multibillion-dollar rounds without ever ringing the opening bell — and Databricks sits at the top of that list. Thirteen years in and still private, the company has now raised roughly $25 billion in equity.

Databricks announced on August 13 that it crossed a $7 billion revenue run-rate in Q2, growing more than 80% year over year, and closed a $5 billion strategic funding round at a $190 billion valuation.

One platform for scattered data and AI

Databricks unifies enterprise data and AI on a single platform and gives teams control over what AI costs and who can access it. Three products anchor the strategy: Lakebase, a serverless Postgres database built for AI agents; Genie, an AI coworker that turns business data into trusted answers and actions; and Unity AI Gateway, which handles governance and cost controls across multiple AI models. More than 20,000 organizations — adidas, AT&T, Bayer, Mastercard, Rivian, Unilever, and 70% of the Fortune 500 among them — run on the platform.

Co-founder and CEO Ali Ghodsi earned his PhD at Sweden’s KTH Royal Institute of Technology before joining UC Berkeley’s AMPLab in 2009, where he worked on Apache Spark. When Spark went open source in 2013, he and six colleagues — including Ion Stoica and Matei Zaharia, the so-called Apache Spark Seven — founded Databricks to commercialize it. He has been CEO since 2016.

The numbers behind the round

• More than 80% year-over-year growth, past a $7B revenue run-rate
• Positive adjusted free cash flow over the trailing 12 months
• Lakehouse, its data warehousing product, past a $1.5B run-rate, growing over 100% YoY
• Lakebase past a $100M run-rate
• Over 1,000 customers at more than $1M run-rate; over 100 at more than $10M

Ghodsi said enterprises no longer want AI that merely talks — they want agents that remember context, deliver accurate answers, and execute work without burning through budgets, and he framed the investor demand for this round as proof the multi-AI strategy is landing. In a separate interview he described demand as running well beyond what the company expected, adding that runaway token spending has rattled enterprise CFOs, which is exactly what drives Gateway adoption.

Coatue led the round, joined by Blackstone, MGX, accounts advised by T. Rowe Price Associates and T. Rowe Price Investment Management, and new investor Sixth Street Growth. BOND, Clearlake Capital, Point72, Premji Invest, and TPG also came in new, alongside existing backers including Andreessen Horowitz, Dragoneer, Fidelity, Franklin Templeton, GIC, Goldman Sachs Alternatives, Insight Partners, J.P. Morgan Private Capital, Morgan Stanley Investment Management, NEA, Ontario Teachers’ Pension Plan, Temasek, Thrive Capital, and WCM.

Coatue co-founder Thomas Laffont said Databricks spent a decade being early to where AI was headed and is now the infrastructure the industry builds on, singling out a pace that compresses multi-year R&D timelines into months — “more like a research lab than a typical software company.”

So how much has it raised while private?

About $25 billion in equity since founding, and north of $32 billion once credit facilities and debt are counted.

The pace is the real story. The $62 billion mark of December 2024 doubled to $134 billion within a year, then climbed another 42% to $190 billion in six months. The term sheet signed in July at $188 billion closed a month later $2 billion higher.

As for an IPO, Ghodsi says the intent is there but the timing is not, pointing to too many distractions in the current market. With Anthropic and OpenAI having already filed confidentially, he said Databricks going public before either of them is very unlikely. Staying private means no quarterly earnings pressure, faster M&A, and valuation upside that accrues to existing shareholders — and Databricks has used that latitude, buying MosaicML for $1.3 billion in 2023, Tabular for over $1 billion in 2024, and security operations platform Panther this past June.

The competition

Snowflake remains the closest public-market comparison, and Databricks’ new valuation sits well above Snowflake’s market cap. Snowflake is pushing into AI on its own terms, having signed a $200 million strategic partnership with Anthropic late last year, and both companies are converging on open data formats. Palantir competes for the same enterprise AI budgets with its blend of data analytics and AI tooling, while Oracle and SingleStore defend the database ground Lakebase is moving into. Databricks’ string of security acquisitions is widening the front further still.