Free
Free plan access to Workers AI, subject to platform limits documented on the limits page.
- Run AI models on serverless GPUs
- Access to 50+ open-source models
- Subject to Workers AI limits
AI API Tool

PRODUCT OVERVIEW
Cloudflare Workers AI is a serverless inference platform that lets developers run machine learning models on GPUs across Cloudflare's global network without managing or scaling infrastructure. It offers a curated catalog of 50+ open-source models for tasks such as text generation, image classification, and object detection, invokable from Workers, Pages, or anywhere via the Cloudflare REST API. A pay-for-what-you-use pricing model means you only pay for consumed resources, and the service is available on Free and Paid plans. It integrates with the broader developer platform, including AI Gateway for caching and rate limiting, Vectorize for vector search, and storage products like R2, D1, and KV. Developers can also use OpenAI-compatible endpoints, function calling, JSON mode, fine-tuning with LoRA adapters, and batch APIs to build AI applications.
Follow the getting started guide to build and deploy your first Workers AI application using Workers bindings or the REST API.
Explore the curated set of 50+ open-source models for tasks like text generation, image classification, and object detection.
Call models running on Cloudflare GPUs from Workers, Pages, or anywhere via the Cloudflare API.
Use OpenAI-compatible API endpoints to integrate Workers AI with existing tooling and SDKs.
Combine Workers AI with AI Gateway for caching and rate limiting, and Vectorize for semantic search and LLM context.
Build AI applications with serverless GPU inference
Run text generation, image classification, and object detection models
Integrate AI models into Workers and Pages applications
Add semantic search and recommendations with Vectorize
Use AI Gateway for caching, rate limiting, and model fallback
It is a serverless platform for running machine learning models on GPUs across Cloudflare's global network, invokable from Workers, Pages, or the Cloudflare API.
Workers AI provides access to 50+ open-source models in its model catalog, covering tasks such as text generation, image classification, and object detection.
It uses a serverless, pay-for-what-you-use pricing model and is available on Free and Paid plans. Specific rates are listed on the pricing page.
You can invoke models from Workers, Pages, or anywhere via the Cloudflare REST API.
Yes, Workers AI documents OpenAI-compatible API endpoints for integration with existing tooling.
Workers AI supports fine-tuning with LoRA adapters, including public LoRA adapters, as documented in its features.
It integrates with AI Gateway for caching and rate limiting, Vectorize for vector search, and storage products like R2, D1, Durable Objects, and KV.
You can visit Cloudflare Support, join the #workers-ai Discord channel, or report Wrangler issues on the Wrangler GitHub repository.
Workers AI is available on Free and Paid plans with a serverless, pay-for-what-you-use pricing model. Specific per-model rates are listed on the pricing page.
Free plan access to Workers AI, subject to platform limits documented on the limits page.
Paid plan with pay-for-what-you-use pricing for Workers AI inference.
COMMUNITY REVIEWS
No reviews yet.