Pruna AI

Pruna AI works on the efficiency of artificial intelligence. The company helps developers and organisations build AI solutions that are smaller, cheaper, faster and more energy-efficient, chiefly through model compression and performance optimisation. Its technical work spans AI efficiency, model compression, AI reliability and sustainability, high-performance model optimisation and open-source machine-learning tooling.

Two offerings anchor the business. The first is an open-source framework for compressing AI models, intended to reduce their size and energy consumption while retaining performance. The second is a line of compressed high-performance models, designed to let builders work with capable AI without trading off speed or cost. The company operates in the AI/ML development, foundation-model and generative-AI, and wider technology sectors.

The team reports decades of combined research and implementation experience in AI efficiency, reliability and sustainability. Pruna AI describes itself as mission-driven, with a stated aim of making AI more efficient and of reducing reliance on Big Tech by widening access to capable models.

Working practices are collaborative and open-source in orientation: the company works hand-in-hand with developers and organisations, and places emphasis on sustainability and on the accessibility of its tooling.