Cloudflare Workers AI
AI & ML Free daily allocation (no card)
Documentation: developers.cloudflare.com ยท Checked Sep 21, 2026
About
Workers AI puts model inference on Cloudflare's edge network: one account gives you 10,000 Neurons a day (about eleven cents of inference), resetting daily, with no card. Calls fail instead of billing past the cap. You get LLMs like Llama, Qwen, and Mistral plus embeddings, image, and audio models. Note a few premium models (Kimi K2.6, GLM-5.x, DeepSeek V4) need a paid billing method; the free lane stays free.
Details
Free to start. You may need a free key, but no credit card.
Auth: Free daily allocation (no card)
Rate limit: 10,000 Neurons/day (~$0.11/day of inference), resets daily at 00:00 UTC; calls fail instead of billing past the cap. Premium models (Kimi K2.6, GLM-5.x, DeepSeek V4) need a paid billing method.
Verified: 2026-09-21
Best for: developer
Setup
- Create a free Cloudflare account (no card).
- Copy your account ID from the dashboard and create an API token.
- POST to the AI run endpoint for the model you want.
- Watch your free allocation in the Workers AI dashboard.
POST https://api.cloudflare.com/client/v4/accounts/{account_id}/ai/run/@cf/meta/llama-3.3-70b-instructUse cases
- Add LLM inference to a serverless app
- Run free embeddings for a RAG agent
- Generate images on the edge for free