Cloudflare AI Gateway is a unified API gateway for AI applications that provides observability, caching, rate limiting, and cost tracking across multiple LLM providers. Available on all Cloudflare plans, the core gateway features are free with no per-call fees beyond the Cloudflare subscription. The platform connects popular providers like Workers AI, Hugging Face, OpenAI, and Anthropic with a single line of code, offering centralized visibility and control. Built into Cloudflare global network infrastructure, AI Gateway provides edge-level caching, request retries, model fallbacks, and analytics. The free tier includes 100,000 AI Gateway logs per month, while the Workers Paid plan starting at USD 5/month provides 1 million logs. In 2026, Cloudflare introduced Unified Billing, allowing customers to pay for third-party model usage directly through Cloudflare invoices. While the platform excels at cost-effectiveness and integration with Cloudflare existing services, it adds 10-50ms of proxy latency, lacks deep AI observability features like token-level tracing, and enforces strict log retention caps that can require manual management at scale.
Free trial available
Cloudflare users who want to add AI gateway capabilities to their existing edge infrastructure
Cloudflare AI Gateway and Respan provide complementary AI infrastructure. Route requests through Cloudflare while gaining deeper observability with Respan.
Top companies in LLM Gateways you can use instead of Cloudflare AI Gateway.
Companies from adjacent layers in the AI stack that work well with Cloudflare AI Gateway.
Last verified: March 9, 2026