Living Models, a startup based in Paris and Berkeley, has announced $7 million in seed funding to develop foundation models for biology. The company focuses on using data from DNA, RNA, and other multi-omics sources to improve the understanding of biological systems.

Training on NVIDIA B200 Cluster

To support the next stage of development, Living Models has secured access to a computing cluster of 120 NVIDIA B200 GPUs. This cluster will be used to train the next generation of biological AI models.

Focus on Agriculture and BOTANIC

Living Models is focusing its efforts on applying AI to biological data, particularly in the agricultural sector. The company introduced BOTANIC, a family of transformer models designed specifically for plant biology. These models are trained on genomic sequences from multiple plant species and analyze genomic and biological data to identify genetic markers associated with traits such as climate resilience and disease resistance.

Accelerating the development of new crop varieties

The main goal is to accelerate the development of new crop varieties by predicting which genetic variants are worth testing. This approach aims to reduce the time required for traditional crop breeding cycles, allowing researchers to focus on the most relevant genetic variants before conducting field validation.

Future Expansion

In the long term, Living Models plans to expand its work on foundation models for biological systems beyond the plant kingdom. The decision to start with plant biology was driven by the availability of large genomic datasets, faster validation cycles compared to other life science fields, and the growing need for technologies that support climate-resilient agriculture.

For those evaluating on-premise deployments, there are trade-offs to consider. AI-RADAR offers analytical frameworks on /llm-onpremise to evaluate these options.