Mirai, a startup focused on optimizing artificial intelligence to run directly on consumer hardware, has secured a $10 million seed round led by Uncork Capital. Founded by Dima Shvets and Alexey Moiseenkov, the company is developing software that improves how AI models perform on smartphones and laptops without relying heavily on cloud infrastructure.
Mirai has built an inference engine optimized for Apple Silicon that increases generation speed while preserving model quality, alongside a developer toolkit designed for rapid integration into applications. The platform targets text and voice use cases initially, with plans to expand into vision and Android devices. The company is also working with model providers and chipmakers to enable efficient edge deployment and hybrid cloud-device orchestration. The funding will support product development as demand grows for cost-efficient, privacy-focused AI computing at the edge.
Featured image: Credit: Mirai
