Cloudflare Workers AI

Cloudflare Workers AI is a serverless platform designed to run AI models. It eliminates the need for managing, scaling, or paying for unused infrastructure. The platform allows for the invocation of models operating on GPUs within Cloudflare's network from your own code. This can be done from Workers, Pages, or through the Cloudflare API.

The platform provides access to over 50 open-source models as part of its model catalog. It operates on a serverless, pay-as-you-use pricing model. Furthermore, it is part of a comprehensive developer platform that includes AI Gateway, Vectorize, Workers, and more.

Workers AI offers a range of features. It includes a selection of popular open-source models for tasks such as image classification, text generation, and object detection. The AI Gateway enables the observation and control of your AI applications with features like caching, rate limiting, request retries, and model fallback. With Vectorize, you can build full-stack AI applications and perform tasks like semantic search, recommendations, and anomaly detection. Workers allows you to build serverless applications and deploy them globally for superior performance, reliability, and scalability. Pages lets you create full-stack applications that are instantly deployed to the Cloudflare global network. Other features include R2 for storing large amounts of unstructured data, D1 for creating new serverless SQL databases for your Workers and Pages projects, Durable Objects for globally distributed coordination API with strongly consistent storage, and KV for creating a global, low-latency, key-value data storage.