- vendor
- Cloudflare
- whatItIs
- Edge-native AI Gateway on Cloudflare's CDN that routes, caches (up to 90% cost reduction), and observes LLM traffic with zero-infrastructure setup.
- hosting
- managed-only
- pricingModel
- included on all Cloudflare plans; 2026 unified 3P billing
- modelProviders
- Workers AI + Anthropic + Gemini + OpenAI + Replicate + Bedrock + Azure
- stateModel
- stateless
- toolModel
- REST API
- includesBrowser
- false
- includesMemory
- false
- longRunning
- per-request
- hitlSupport
- false
- observability
- built-in analytics + identity-aware usage tracking
- maturity
- ga
- launched
- 2024
- notes
- Automatic caching up to 90% cost savings; model fallback + rate-limiting built-in. Aug 2026: unified billing with Workers AI (single control plane, prepaid credits work for both); Identity-Aware AI Gateway (open beta) attaches verified identity via Cloudflare Access to every AI request, with per-user/agent spend limits and anomalous-usage flagging. Jun 2026: added cost-based spend limits by model/provider with configurable time windows.
