What is Cloudflare Workers AI
Cloudflare Workers AI provides serverless model inference on Cloudflare's global developer platform, integrated with Workers, AI Gateway and Vectorize.
Free allocation
As of September 17, 2026, Workers AI includes 10,000 Neurons per day at no charge on both Workers Free and Workers Paid. Free accounts stop accepting further AI operations after the daily allocation is exhausted. Current published rates include Llama 3.2 1B Instruct at US$0.027 per million input tokens and US$0.201 per million output tokens, GPT OSS 120B at US$0.35 and US$0.75, and GLM 5.3 Flash at US$0.15 and US$0.50.
Workers Paid
Usage beyond the free daily allocation requires Workers Paid, which has a US$5 monthly account minimum. Additional Workers AI usage costs US$0.011 per 1,000 Neurons, with per-model token or unit prices expressed as equivalent Neuron consumption. The broader paid Workers plan includes 10 million requests and 30 million CPU milliseconds monthly, then US$0.30 per additional million requests and US$0.02 per additional million CPU milliseconds.
Workers AI uses serverless scaling and per-model pricing across language, image, speech, embedding and other tasks. Users can monitor Neuron use in the dashboard; daily allocations reset at 00:00 UTC.
Official sources: Workers AI pricing, Workers platform pricing, Workers AI documentation.