Shumai — High-Speed TypeScript Tensor Library

Shumai is an open-source tensor library for building fast deep learning applications in TypeScript and JavaScript.

Shumai is an open-source tensor library designed to bring high-performance deep learning capabilities to the TypeScript and JavaScript ecosystem. Developed by Meta AI Research, it provides a bridge for web developers and software engineers who need to execute complex mathematical operations and machine learning models without leaving their primary programming environment. By utilizing Bun and a C++ backend, it ensures that data-heavy computations reach speeds comparable to Python-based frameworks while maintaining a familiar syntax for front-end and full-stack professionals. The library allows business owners to modernize their web-based products by integrating differentiable programming directly into their technical stack. Shumai enables the creation of network-connected applications that can distribute tensor operations across multiple nodes, making it a suitable choice for scalable infrastructure. On surfai.app, we track tools like Shumai that help technical teams reduce the overhead of managing legacy Python microservices for simple inference or training tasks. Key features include automatic differentiation, support for large-scale gradients, and a design philosophy that prioritizes low-latency execution. For businesses building intelligent web applications, this tool simplifies the deployment of neural networks and custom optimization algorithms. Shumai is entirely open-source and free to use for any commercial application under its current license.

• Pricing: free

Specifications

Other tools in this category

• LaunchDarkly
• PullRequest
• SQL CREATOR

Compare tools in this category

Visit Shumai (Meta)