Databricks funding: Every round, investor, and valuation (2026)

Databricks sells a data platform that companies use to store, clean, and analyze their data, and increasingly to build AI agents on top of it. It was founded in 2013 by the UC Berkeley researchers who created Apache Spark, and it charges customers for the computing they use rather than a flat license. More than 20,000 organizations now run on it, including about 70% of the Fortune 500.

Databricks has raised about $25 billion in equity across 13 rounds, plus roughly $7 billion in credit lines. Its latest round, closed in August 2026, valued the company at $190 billion.

Early money came from two venture firms. Then the cloud giants that resell Databricks joined, and the recent rounds came from sovereign wealth funds and asset managers. Let's dive into Databricks' funding history.

Databricks' funding falls into three eras: a venture-backed open-source startup, a growth company courted by the cloud providers, and an AI platform funded like a pre-IPO giant. Each era began when revenue reached a new tier.

Andreessen Horowitz backed Databricks in 2013, before it had a product. The founders had built Spark as a free research project, and the plan was to sell a hosted cloud version of it to companies that did not want to run it themselves.

NEA led the next two rounds, and the hosted product launched alongside the Series B in 2014. In between, In-Q-Tel, the venture firm backed by the US intelligence community, invested to adapt the platform for US intelligence agencies. Andreessen Horowitz returned to lead the 2017 Series D, which brought total funding to about $247 million.

Microsoft first invested in the 2019 Series E. By then, Databricks had passed $100 million in annual recurring revenue, and Microsoft was already selling it as Azure Databricks.

Amazon Web Services, Alphabet's growth fund CapitalG, and Salesforce Ventures all joined the 2021 Series G, while Microsoft invested again. Six months later, Morgan Stanley's Counterpoint Global led the Series H, and the three clouds came back for its final close.

The Series I in 2023 arrived during a slump in late-stage funding, yet it still raised the price. Nvidia joined as a strategic investor, and Qatar Investment Authority came in through a later close.

The Series J was the largest round Databricks had raised by a factor of six. Thrive Capital led it, and Databricks called the financing non-dilutive: part of the money bought shares from current and former employees and paid their related taxes. The Stack reported that most of the first $8.6 billion was a secondary share sale, though Databricks never gave the split.

JPMorgan arranged Databricks' first credit package in January 2025, split between a revolving credit line and a term loan. A year later, an extension lifted its total borrowing capacity to about $7 billion.

The Series K was small by comparison and came from existing investors. The round was already oversubscribed when Databricks announced it in August 2025, and it closed the following month.

The Series L opened in December 2025 at more than $4 billion and grew to about $5 billion by the February close. Coatue then signed a term sheet in July 2026 at $188 billion, and the round closed the next month with Blackstone, MGX, T. Rowe Price, and first-time investor Sixth Street Growth alongside it.

Databricks' valuation rose about 69-fold between the 2019 Series E and the August 2026 round. Most of that gain came in two bursts: 2021, and the 20 months after the Series J.

The 2021 jump came when software stocks were near their peak. The Series I then held the price roughly flat through the 2022 downturn, when many late-stage startups raised at lower valuations or stayed away from the market.

Databricks expected to cross a $3 billion run rate when it announced the Series J, and it passed $7 billion in the quarter that ended in July 2026, growing more than 80% a year. That puts the latest price at roughly 27 times annualized revenue, up from about 21 times at the Series J.

The latest price also tops the market value of Snowflake, its closest rival, and how Snowflake is owned as a listed company makes a useful contrast. Each round has diluted the seven co-founders and early funds, which changes how Databricks' ownership is split.

Databricks' investors fall into four groups: the early venture firms, the growth and crossover funds, the tech companies that sell or use its software, and the sovereign and institutional money behind the latest rounds.

Andreessen Horowitz led the Series A and three more rounds through 2019, then co-led the Series J and the Series K. It also invested in the Series L and the August 2026 round, which makes it the most consistent backer on the cap table.

NEA led the Series B and Series C and still participates, most recently in the August 2026 round. Between them, the two firms led every round in Databricks' first four years.

Thrive Capital, Insight Partners, and WCM co-led both the Series J and the Series K, and Insight also co-led the Series L. Thrive's chief executive, Joshua Kushner, called the firm a long-term partner when it led the Series J.

Coatue has invested since the 2019 Series E and led the August 2026 round. Other repeat names include T. Rowe Price, Fidelity, Franklin Templeton, BlackRock, Morgan Stanley's Counterpoint Global, Dragoneer, and Ontario Teachers' Pension Plan.

Microsoft sells Databricks as a native Azure service and has invested since 2019, including in the Series L. It also competes with Databricks through its own Fabric data platform, so a company with Microsoft's widely held ownership is backing a private rival to one of its own products.

AWS and CapitalG invested in the 2021 Series G, so all three major clouds are shareholders. Salesforce Ventures joined in the same round.

Nvidia joined the Series I in 2023, three months after Databricks bought the AI model startup MosaicML. Its stake ties into Nvidia's chip-selling business, since Databricks' AI tools run on the GPUs Nvidia sells to the clouds.

Meta joined the Series J as a strategic investor. Databricks hosts Meta's open Llama models on its platform. The investment sits on the balance sheet of a company where Meta's founder-controlled ownership gives Mark Zuckerberg the final say.

GIC has invested since 2021 and co-led the Series J. Qatar Investment Authority joined the Series I, and Temasek and Abu Dhabi's MGX came in with the Series J. MGX co-led the Series K and invested again in August 2026.

J.P. Morgan Asset Management co-led the Series L. Blackstone, Goldman Sachs, and Point72 are among the Wall Street asset managers in the latest round.

Databricks has reported positive free cash flow since late 2024, so its funding is not paying for losses. The money goes to employee liquidity, acquisitions, AI products, and international expansion.

Part of the Series J went to employees rather than the balance sheet. Staff paid in stock units owe tax as those shares vest, and a private company offers no market to sell them in.

Databricks listed employee liquidity again as a use of the Series L money. Ghodsi has said the main reason to eventually go public is to give employees a market for their shares.

Databricks paid about $1.3 billion for MosaicML in 2023, including retention packages, and more than $1 billion for Tabular in 2024. Neon, a serverless Postgres database startup, cost about $1 billion in 2025 and became the base for Lakebase.

In 2026, Databricks has bought Quotient AI, SiftD, Panther, Electric, and Row Zero. Ghodsi has said more deals are coming.

The recent rounds fund Databricks' push from data storage into AI applications. Its main products there are Agent Bricks for building AI agents, Lakebase as an operational database for those agents, and Genie, an assistant that answers business questions in plain language.

Databricks also pays for access to frontier models. It signed $100 million partnerships with Anthropic in March 2025 and OpenAI in September 2025, bringing their models into its platform. Those deals connect Databricks to Anthropic's enterprise model business and to how OpenAI sells to enterprises.

Databricks opened regional hubs in London and Singapore in 2024 and expanded in Latin America and the Middle East. With the Series L, it said it would add thousands of jobs in Asia, Europe, and Latin America.

It also signed new office leases in San Francisco and Sunnyvale to recruit AI researchers. The spending ties back to how Databricks makes money: every new customer workload adds usage, and usage is what Databricks bills.

Databricks' Lakehouse data warehousing business passed a $1.5 billion run rate in mid-2026, more than doubling in a year, and Lakebase, launched in 2025, passed a $100 million run rate. These newer products are growing fastest, and Databricks is pushing further into large enterprises and into regions outside the US.

Databricks launched Lakewatch, a security data product powered by Anthropic's Claude models, in March 2026, then agreed to buy Panther to strengthen it, while the Row Zero deal adds spreadsheets to Genie. Both deals widen the product line from data into security and business apps. The August round is paying for this push, along with the Unity AI Gateway, which helps customers control which AI models they use and what they spend on them.

Ghodsi told Bloomberg in June 2026 that this was "a terrible year to go public," and in August he told CNBC the company still plans to list once the public market is less of a distraction, which points to 2027 at the earliest. The main risks are competition from Snowflake and from the same cloud providers that resell Databricks, customers cutting AI spending as model costs rise, and a valuation that assumes growth near 80% continues. OpenAI's own funding history shows how much capital is chasing the same enterprise AI budgets.

Databricks has raised about $25 billion in equity across 13 rounds since 2013. The largest were the $10 billion Series J in 2024 and 2025, the roughly $5 billion Series L in 2025 and 2026, and the $5 billion round in August 2026. It also has about $7 billion in debt facilities, which brings the total to about $32 billion.

Databricks' investors include Andreessen Horowitz, NEA, Thrive Capital, Insight Partners, Coatue, T. Rowe Price, Fidelity, and Franklin Templeton. Strategic investors include Microsoft, AWS, CapitalG, Salesforce Ventures, Nvidia, and Meta, and sovereign funds include GIC, QIA, Temasek, and MGX.

Databricks was valued at $190 billion after its $5 billion round closed on August 13, 2026. That was up from $134 billion in the Series L six months earlier and $62 billion in the Series J.

No. Databricks is private and has not filed for an IPO. Ali Ghodsi has said the company will go public but called 2026 a bad year to list, which points to 2027 at the earliest.

Databricks is owned by its seven co-founders, employees, and its investors. Its longest-standing outside backers are Andreessen Horowitz and NEA, and the company does not disclose individual stakes. CEO Ali Ghodsi and executive chairman Ion Stoica lead the company.