On Thursday, Databricks announced a new funding round that values the company 188 billion dollars. The tour was led by Coatue.
Databricks didn’t say exactly how much it raised; said the money is not yet in its hands and the round will close later this summer. (Other publications have since reported an increase of approx 3 billion dollars.) While it’s unusual for a company to announce the money before it receives it, the VC tells TechCrunch the deal was solid, with so many firms wanting it that there was no reason to keep the company’s shiny new valuation a secret.
In fact, Databricks has raised money for a year and a half as it successfully transformed its image from just a former SaaS sensation to an AI provider. Going back to yesterday BC (before ChatGPT).
Just five months ago, in February, Databricks closed a $5 billion Series L raise It was valued at 134 billion dollars. Five months before that, in September 2025 It raised $1 billion at a cost of $100 billion. And about nine months before that, in December 2024, he raised the a It broke a record worth of 10 billion dollars It is worth 62 billion dollars.
Databricks has raised so many rounds over the years that it’s the subject of this last one memories of running out of letters of the alphabet. “We trigger alerts when we get the AA series,” one person wrote.
However, the reconstruction of his image has been legal. Founded in 2013, it initially found success in the big data era with software that allowed businesses to store large amounts of data in the cloud while developing rapid analytics.
Because it already sits on a treasure trove of corporate data, Databricks is well-positioned to respond when companies start asking for AI with the same security and governance they expect from traditional enterprise software.
The company began to roll out artificial intelligence products one after another Lakebase, a database built for AI agentsand Unity, along with its AI gateway and “meta-harness” called Omnigent, which manages multiple agents.
Databricks is also growing it became known As one of the great examples of enterprises adopting Chinese-based open weight models (models where the underlying code is published for anyone to use and modify) for cost management, It is one of the biggest trends of 2026. He is a particular champion of Z.ai’s GLM 5.2 as a model for coding.
Last week Databricks CEO Ali Ghodsi shared the results Some internal benchmarks for 3,000 software engineers to manage their AI spend.
The company compared its AI models with actual tasks performed by its programmers. It is not surprising that in a blog post announcing the resultsDatabricks shared that “open models, and GLM 5.2 in particular, can now handle the most advanced coding tasks” and at a lower cost than Anthropic and OpenAI’s proprietary models.
But it surprised people that the choice of harness — an agent coding tool like Codex or Claude Code that encapsulates a model and manages its context and instructions — affected costs just as much. He found that the open source Pi harness is one of the best at handling the context surrounding each command, and therefore one of the lowest cost options without sacrificing quality.
“The lesson here is not that a trailer is always cheaper or that local trailers are worse” announced the post. “Instead, model selection is only one piece of the puzzle.”
All of this added to Databricks’ image as an AI company, even if it wasn’t founded as an AI lab. This in turn gave him an AI-halo to raise money and raise his value. As we reported earlier, the AI effect is so strong these days that even noted sandwich shop Jersey Mike’s AI 22 times in S-1 documents.
When you purchase through links in our articles, we may earn a small commission. This does not affect our editorial independence.
