Databricks raises $5B at $190B, and the multiple barely moved

Databricks has raised another $5 billion, led by Coatue with Blackstone, MGX, T. Rowe Price and new investor Sixth Street Growth, at a $190 billion valuation ^[4]. The company says it has passed a $7 billion revenue run rate with more than 80% year-over-year growth in Q2 ^[5], which prices the business at roughly 27 times current revenue ^[19] and makes the entry price a bet on persistence rather than mere continuation.

The comparison worth making is not to software comps but to Databricks eight months ago. In December 2025 it was valued at $134 billion on a run rate it had just said exceeded $4.8 billion ^[24], about 28 times ^[25]. Since then the valuation is up about 42% and the run rate about 46% ^[20]. Investors did not re-rate the multiple; they paid for revenue that arrived, and the multiple compressed slightly. That is a more disciplined print than the headline number suggests.

The growth curve is the actual asset. Databricks went from 50% to 80% growth over four quarters, thirty points of acceleration between $4 billion and $7 billion of run rate ^[8]. Per SaaStr's read of the release, this quarter held at 80% rather than climbing, which makes the acceleration a completed arc rather than something still in motion ^[9]. There is also a measurement problem: ">$7 billion" is a threshold, not a figure, 80% growth on last year's $4.0 billion implies at least $7.2 billion, and the roughly $1.5 billion of run rate added between the January and April quarters looks much larger than the step from April's $6.9 billion to July's disclosure ^[10]. The number is self-reported by a private company with no auditor and no obligation to define it consistently ^[6].

Run the forward arithmetic and the case gets clearer. Hold 80% for a year and the run rate is about $12.6 billion, or 15 times today's price ^[21]. Halve growth to 40% and it is about $9.8 billion, or 19 times ^[22]. Both are survivable. Neither leaves room for the deceleration that companies at this scale normally deliver.

The margin line is where the model gets tested. CEO Ali Ghodsi told CNBC in June that margins are shrinking because agents generate far more queries than humans do ^[11]. Databricks prices consumption, so it captures that traffic directly, and is still absorbing compression. Vendors on flat per-seat pricing are absorbing worse: Figma gave up five points of gross margin year over year to AI credits ^[1], Atlassian guided non-GAAP operating margin from 36% down to 25% for FY27 ^[2], and Canva cut its growth forecast by a third and started metering Pro features ^[3].

Revenue quality reads better than the valuation. More than 1,000 customers now consume above $1 million of run rate, up from 650-plus in September 2025, roughly 54% growth in that cohort in eleven months, and more than 100 consume above $10 million ^[12][13]. Lakebase passed a $100 million run rate from a standing start ^[17]. Databricks passed Snowflake at the company level in the October 2025 quarter ^[14], though Snowflake still adds more absolute product dollars, roughly $1.8 billion a year on 34% growth and 126% net revenue retention, than Lakehouse's roughly $1.5 billion ^[15][16].

Watch the sequential add next quarter, because that is where a threshold disclosure stops hiding a slowdown ^[10]. Watch the $1 million cohort count ^[12], and whether gross margin gets disclosed at all. And note the funding environment around this: River AI, founded this year, raised $1.1 billion across seed and Series A with Nvidia and AMD Ventures participating ^[18]. Databricks has raised roughly $25 billion since a $13 million Series A in 2013, about 3.6 times its current run rate ^[7][23].