Organisation
Databricks
Data and analytics platform, built by the creators of Apache Spark, that helps businesses train and fine-tune their own models on their own data.
Databricks is a data and analytics company, founded in 2013 by the creators of the open-source Apache Spark project and led by Ali Ghodsi, that sells a platform for storing and processing the large datasets companies use to build software and, increasingly, AI. Rather than compete as a frontier lab, it has positioned itself around helping businesses train and fine-tune their own models on their own data: it bought the model-training start-up MosaicML for $1.3 billion in 2023 and released the open-weight DBRX model the following year. It has raised money at some of the highest valuations of any private AI-adjacent company, reaching a reported $134 billion in late 2025, on the bet that open, customisable models are a durable commercial foundation and not just a research artefact.
- Category
- Data, training & services
- Founded
- 2013
- HQ
- San Francisco, US
- Funding
- Reported ~$134bn valuation (2025)
- Key people
- Ali Ghodsi
Appears alongside
Featured in threads
Tracks
- Open weights & ecosystem 4
- Money & business 3
- Models & capabilities 3
Databricks raises at $134bn valuation on $4.8bn revenue run-rate
The Series L, led by Insight Partners, Fidelity and JPMorgan Asset Management, valued the data and AI infrastructure company 34% above its round four months earlier.
Money & business
Databricks raises $10bn Series J at $62bn valuation
Led by Thrive Capital with participation from Andreessen Horowitz, GIC and others, the data-and-AI platform company said it expected to cross $3bn in annualised revenue.
Money & business
Databricks releases DBRX
Built by the former MosaicML team Databricks had acquired the previous year, the 132-billion-parameter model activated only 36 billion parameters per input.
Open weights & ecosystem · Models & capabilities
Databricks agrees to acquire MosaicML for $1.3bn
MosaicML's open MPT-7B model had been downloaded more than 3.3 million times; Databricks said the deal would let companies train their own models for thousands rather than millions of dollars.
Money & business · Open weights & ecosystem
MosaicML releases MPT-7B
Four variants shipped together, including a version extrapolating to roughly 84,000 tokens of context — far beyond the 2,000-4,000 tokens typical of open models at the time.
Open weights & ecosystem · Models & capabilities
Databricks releases Dolly 2.0
Databricks released Dolly 2.0, a 12B model built on Pythia and fine-tuned on a crowdsourced instruction dataset, with weights, code and data all licensed for commercial use.
Open weights & ecosystem · Models & capabilities