Home

ai gateway

Cloudflare AI Gateway

Cloudflare AI Gateway is a control and observability layer for applications that call AI models. It is intended for developers and teams that need request analytics, logging, caching, rate limits, retries, or model fallbacks across supported providers.

Data status: partially_verified · Checked: 2026-10-16

Official website ↗

Features and benefits

  • Request analytics including request counts, tokens, and costs
  • Request logging and error visibility
  • Caching, rate limiting, retries, and model fallback
  • Works with providers including Workers AI, Anthropic, Google Gemini, OpenAI, and Replicate

Limitations

  • Limits differ by feature: legacy log storage is 100,000 logs total on Workers Free and 10,000,000 per gateway on Workers Paid; newer accounts follow Workers Logs pricing and retention.
  • Guardrails inference and Logpush exports can incur separate usage-based charges.
  • Unified Billing adds a 5% fee to purchased credits; provider inference rates are passed through without markup.

Pricing

Core features

0 USD · free

Dashboard analytics, caching, and rate limiting are free on all plans; requires a Cloudflare account. Workers Logs, legacy log storage, Guardrails inference, Logpush exports, and provider inference have separate limits or usage charges.

Pricing source

Workers AI inference via Unified Billing

Price unknown · usage_based

Per-model token/compute pricing; 5% fee applies to purchased credits, while provider inference rates pass through without markup. See current model-specific rates at source.

Pricing source

Sources