Databricks Hits $188B Valuation as AI Fundraising Run Continues
Databricks announced a new funding round valuing it at $188B, led by Coatue. Here's what drove the leap from $134B in just five months.
Databricks announced on July 17, 2026 that it is raising a new funding round led by Coatue at a $188 billion valuation. The company has not confirmed the exact size, but the round is reported to be around $3 billion and is expected to close later this summer. The announcement came before the money even landed, which is unusual but reflects strong investor demand. The raise continues an 18-month streak that has taken Databricks from a $62 billion valuation to $188 billion, fueled by its pivot from data infrastructure into enterprise AI.
What happened
| Date | Valuation / Round |
|---|---|
| December 2024 | $62B (record $10B raise at the time) |
| September 2025 | $100B ($1B raised) |
| February 2026 | $134B (Series L, $5B raised) |
| July 17, 2026 | $188B (~$3B raise, closing later in summer) |
Databricks made its latest valuation public before the round formally closed, which is not standard practice. According to a venture capitalist who spoke to TechCrunch, investor demand was strong enough that the company had little reason to stay quiet. The round is led by Coatue, with many firms reportedly seeking in. The exact figure has not been confirmed by Databricks, but reporting from other outlets puts it at roughly $3 billion.
The pace of the raises has become something of a running joke online. Because the company has raised so many rounds, observers noted it is running through the alphabet of series letters. “Turning on alerts for when we get a Series AA,” one commenter posted.
How Databricks built an AI story from a data company
Databricks was founded in 2013 and grew through the big data era, selling enterprises cloud software for storing large volumes of data and running fast analytics. That existing position on top of enterprise data turned out to be a useful starting point when companies started wanting AI with the same security and governance they apply to traditional software.
Over the past year the company has rolled out a string of AI-focused products:
- Lakebase: a database designed specifically for AI agents
- Unity: an AI gateway for managing model access and governance
- Omnigent: a “meta-harness” that coordinates multiple AI agents
These products, combined with its reputation for helping enterprises adopt open-weight AI models at lower cost, have repositioned Databricks in the market. The company’s story is now less about data pipelines and more about enterprise-grade AI infrastructure, a framing investors are clearly rewarding.
What does Databricks’ own AI cost research show?
Last week, CEO Ali Ghodsi published results from internal benchmarking the company ran to control AI spending across its 3,000 software engineers. The team tested AI models on the actual coding tasks its programmers perform, not abstract benchmarks.
The findings were notable on two fronts. First, Databricks concluded that Z.ai’s open-weight model GLM 5.2 could handle even the most complex coding tasks at a lower total cost than proprietary models from Anthropic and OpenAI. Second, and more surprisingly, the choice of “harness” (the agentic coding tool that wraps around a model and manages its context and instructions, such as Codex or Claude Code) had an equally significant effect on cost. The open-source harness Pi was identified as one of the lowest-cost options without a quality penalty.
The company’s own conclusion: “model choice is only one piece of the puzzle.”
This kind of public, practical cost research is part of how Databricks has built credibility with enterprise buyers. It positions the company as a practitioner, not just a vendor. Businesses looking to integrate AI into their workflows are asking exactly these questions about cost versus capability tradeoffs.
Why it matters
Databricks’ trajectory from $62B to $188B in 18 months is a clean case study in what the AI halo looks like in practice. The company did not start as an AI lab. It started as data infrastructure. Its leap in valuation comes from combining real enterprise data assets with a credible AI product roadmap, not from building foundation models.
For business operators, the more relevant story is the benchmarking. If a company running 3,000 engineers found that an open-weight Chinese model and an open-source harness beat out Anthropic and OpenAI on cost for coding tasks, that is worth examining for your own AI tooling budget. The context management layer (the harness) is underappreciated as a cost driver.
The broader trend Databricks represents is enterprise AI moving away from “use whatever the biggest model is” toward deliberate cost governance. That shift is only going to accelerate as AI spend becomes a line item CFOs scrutinize. You can follow more coverage of this pattern in our AI news section.
Our take
The valuation number is eye-catching, but the more interesting data point is the benchmarking post. Most companies at this stage are selling the dream. Databricks is publishing specifics: which model, which harness, and why the combination matters for real engineering tasks. That is either genuinely useful research or very well-executed marketing, and the answer is probably both.
The finding that harness choice matters as much as model choice is underreported and practically important. If you are spending serious money on AI coding tools and have not evaluated how context management affects your per-task cost, you are leaving money on the table. The AI effect on valuations is real, but for most businesses the more actionable news is in the cost data, not the funding announcement.
What to do about it
- Audit which AI models and harnesses your engineering or ops team is currently using and at what cost per task.
- Test at least one open-weight model alternative alongside your current proprietary model on a real internal task, not a synthetic benchmark.
- Evaluate harness options separately from model choice. Context management is a distinct cost variable.
- If you want help building a structured approach to AI tooling costs, talk to the Lumien team about what we have seen work for similar-sized businesses.
The companies getting the most value from AI right now are the ones treating it like an ops problem, not a technology bet.
Frequently asked questions
What is Databricks' valuation in 2026?
Databricks announced a $188 billion valuation on July 17, 2026, up from $134 billion in February 2026 and $62 billion in December 2024.
How much did Databricks raise in its latest round?
Databricks has not officially disclosed the amount. Other outlets report the raise is approximately $3 billion. The round was led by Coatue and had not formally closed at the time of the announcement.
What AI products does Databricks offer?
Databricks has launched Lakebase (a database for AI agents), Unity (an AI gateway), and Omnigent (a meta-harness for managing multiple AI agents).
Is GLM 5.2 cheaper than OpenAI or Anthropic models for coding?
According to Databricks' internal benchmarking of tasks performed by its 3,000 engineers, Z.ai's open-weight model GLM 5.2 handled the highest difficulty coding tasks at a lower total cost than proprietary models from OpenAI and Anthropic.
