Databricks Raises $5B at $190B Valuation After Investors Offered $15B

Databricks closed a $5 billion funding round on August 13, 2026, valuing the AI data platform at $190 billion, after investor demand far outstripped what the company had planned. CEO Ali Ghodsi told TechCrunch the company originally wanted to raise just $1 billion, but a news report published mid-conference triggered so much inbound interest that a select group of investors alone expressed $15 billion of appetite. The round was led by Coatue and included Blackstone, MGX, T. Rowe Price accounts, and new investor Sixth Street Growth, with roughly two dozen VCs named as participants.

What happened

Fact	Detail
Round size	$5 billion
Valuation	$190 billion
Close date	August 13, 2026
Lead investor	Coatue
Other key investors	Blackstone, MGX, T. Rowe Price, Sixth Street Growth
Investor demand (from select group)	$15 billion
Original fundraising target	$1 billion
Total raised in past 20 months	$20 billion
Annualized revenue run rate	$7 billion, growing 80% year-over-year
Core data warehouse run rate	$1.5 billion, growing 100% year-over-year
Lakebase (agent database) run rate	$100 million

The round’s origin story is unusual. According to CEO Ali Ghodsi, Databricks was planning a modest $1 billion raise when a report in The Information described an imminent large fundraise. The article ran while Databricks was running its own customer conference in June. Ghodsi described his phone blowing up with investor calls at the worst possible moment. A curated group of investors then signalled $15 billion of willingness to invest, far more than Databricks needed or wanted to issue.

Rather than say no to long-standing backers and risk bad blood, Databricks issued more shares than planned. A July press release confirmed the round had closed at $188 billion. Thursday’s announcement revised the valuation to $190 billion and disclosed the full $5 billion figure for the first time.

Why is Databricks worth $190 billion?

The numbers give a partial answer. At $7 billion in annualized revenue growing 80% year-over-year and cash-flow positive, Databricks is not a story built purely on promises. The cloud data warehouse product at $1.5 billion run rate is still accelerating at 100% annually. Lakebase, a database built for AI agents launched in June 2025, has already reached $100 million in annualized revenue. Genie, an AI tool for on-the-spot business analysis, is growing fast by Ghodsi’s account, though no specific figure was given.

The company also keeps acquiring. This week it announced the purchase of Electric, the company behind PGlite, a lightweight Postgres database that lets AI agents spin up their own databases on demand. In June it bought Panther, an AI cybersecurity company. In March it bought two additional startups. Terms were not disclosed for any of these deals.

Why it matters

For anyone watching the AI infrastructure market, Databricks’ raise illustrates how much institutional capital is chasing a small number of credible bets. The gap between a $1 billion target and $15 billion of demand tells you something about the scarcity of companies with real revenue at this scale. Sixth Street Growth, the firm co-founded by former Goldman Sachs chief investment officer Alan Waxman, joined as a new investor, which suggests the draw extends beyond traditional tech-sector VCs.

The continuing preference to stay private is also notable. Ghodsi told CNBC he wants to take Databricks public eventually, but with multibillion-dollar cloud commitments to all three major hyperscalers and a 100-person AI research team, the burn rate makes a quiet private environment more attractive right now. With $20 billion raised over 20 months, the runway is long.

For businesses evaluating AI data tools, Databricks is becoming harder to ignore as a default infrastructure layer. Lakebase in particular, sitting at $100 million run rate less than 14 months after launch, is worth watching if you are building anything that involves AI agents querying or generating data.

Our take

The fundraising mechanics here are almost a side story. What stands out is the revenue profile: 80% growth at $7 billion annualized is not common at that scale, and 100% growth on the core warehouse product suggests demand is not slowing. The Lakebase number is the one we would focus on. A database product purpose-built for agents hitting nine figures in run rate within its first year is a signal about where enterprise AI budgets are actually going: not into chatbots, but into the data plumbing underneath them.

If you are helping clients integrate AI into their business processes, as we do through our AI integration work, the question of where data lives and how agents access it is increasingly central. The acquisitions also suggest Databricks is trying to own more of the agent infrastructure stack, not just the analytics layer it started with.

On the valuation: $190 billion is a lot. But compared to companies with far weaker fundamentals trading at similar or higher multiples in public markets right now, the number is at least tethered to something real. We cover more on the broader AI funding environment in our AI news section.

What to do about it

1. If you use any cloud data warehouse today, check whether Databricks or its competitors are being evaluated by your data team. The pricing and feature gap is shifting fast.
2. Look at Lakebase if you are building agent workflows that require persistent, queryable data storage. It launched in June 2025 and is already at scale.
3. Track the acquisitions. Electric’s PGlite integration will affect how lightweight Postgres databases are used in agent architectures.
4. Watch for an IPO signal. With this many investors who will need an exit, a public offering is likely within the next two to three years.

The practical takeaway: Databricks is building the data infrastructure layer for the agent era, and the revenue numbers suggest enterprises are already paying for it.

Source: TechCrunch · AI