MOSAICML’S PLATFORM FOR TRAINING LARGE LANGUAGE MODELS AND GENERATIVE AI TOOLS WILL COMPLEMENT DATABRICKS LAKEHOUSE PLATFORM FOR AI AND DATA SERVICES. THE ACQUISITION PROVIDES DATABRICKS’ CLIENTS WITH AN OPPORTUNITY TO DEVELOP GENERATIVE AI PROJECTS WITH THEM RATHER THAN TURNING TO OPENAI OR OTHER MODEL PROVIDERS. DATABRICKS BELIEVES THAT MERGING MOSAICML’S MODELS WITH ITS EXISTING TECHNOLOGY WILL REDUCE THE PRICE FOR TRAINING AND DEPLOYING LLMS FROM MILLIONS TO THOUSANDS OF DOLLARS.
Data storage and analytics company Databricks has acquired generative AI model start-up MosaicML for $1.3bn. With the acquisition, MosaicML’s platform for large language models and generative AI training will become a part of Databricks Lakehouse Platform for AI and data services. MosaicML developed the MPT LLMs, which have different parameter sizes and have become very popular among clients, including the Allen Institute for AI, Hippocratic AI and Replit. Its “automatic optimisation of model training” completes the training process up to seven times faster than the more common methods.
The acquisition has potential for Databricks’ customers further to develop generative AI projects with the company, rather than turning to OpenAI or other model providers. The company claims merging MosaicML’s models with Databricks’ existing tech will drop the price for training and deploying LLMs from millions to thousands of dollars. The entire staff of MosaicML will join Databricks as part of the acquisition. The acquisition price sextuples the $222m valuation for MosaicML, which has raised about $64m in total.
“Every organisation should be able to benefit from the AI revolution with more control over how their data is used,” said Databricks CEO Ali Ghodsi. “Databricks and MosaicML have an incredible opportunity to democratise AI and make the Lakehouse the best place to build generative AI and LLMs. Databricks and MosaicML’s shared vision, rooted in transparency and a history of open source contributions, will deliver value to our customers as they navigate the biggest computing revolution of our time.”