Databricks, the fast-growing data infrastructure company, has announced a partnership with Mistral, a Paris-based startup known for its highly-performant large language models (LLMs). Databricks will be investing in Mistral and integrating select Mistral LLMs into its data intelligence platform. This move will make it easier for enterprise users to utilize these models for generative AI applications without compromising security, privacy, and governance.
Mistral has been making waves in the AI community with its growing family of open-source LLMs. The partnership with Databricks is another significant distribution opportunity for Mistral, following similar partnerships with Snowflake and Microsoft. This has raised concerns about corporate capture and attracted the attention of regulators.
Databricks has confirmed that the partnership will result in the native integration of two text-generation models from Mistral – Mistral 7B and Mixtral 8x7B. Both models are open source and offer impressive capabilities. Mistral 7B is a small transformer model with 7 billion parameters, while Mixtral 8x7B is a sparse mixture of expert models capable of handling multiple languages. In benchmark tests, Mixtral 8x7B outperformed models from Meta and OpenAI.
Users of the Databricks Data Intelligence Platform can now access these models through the platform’s marketplace. They can experiment with the models in the Mosaic AI Playground, use them as optimized model endpoints through Mosaic AI Model Serving, or customize them for specific use cases using proprietary data hosted on the platform.
It’s worth noting that the partnership between Databricks and Mistral does not mention Mistral Large, the latest model from Mistral that outperforms other leading models. Snowflake has already integrated Mistral Large into its data cloud’s Cortex service, but its inclusion in Databricks’ platform remains uncertain.
Mistral has been actively pursuing partnerships to expand its reach and establish itself as a trusted player in the AI category. In addition to Databricks and Snowflake, Mistral has secured partnerships with Microsoft, IBM, Perplexity, and Amazon. These collaborations are key to driving AI use cases across various sectors.
Overall, the partnership between Databricks and Mistral is a significant development in the AI and data intelligence space. It provides Databricks customers with access to powerful LLMs and expands Mistral’s distribution network. As Mistral continues to secure partnerships, it will be interesting to see how its prominence grows and how it contributes to advancing AI technologies.