Beyond basic
AI proxying
Cloudflare AI Gateway is a solid proxy. But proxying is not routing. Requesty gives you intelligent model selection, cross provider failover, deep cost analytics, and enterprise governance. All fully managed.
$ # create gateway in CF dashboard...
$ # get API key from OpenAI...
$ # get API key from Anthropic...
$ # get API key from Google...
$ # configure each provider separately
$ # no smart routing between them
// Separate keys, separate bills, basic proxy
$ export OPENAI_BASE_URL="https://router.requesty.ai/v1"
$ export OPENAI_API_KEY="rq-..."
// One key. 400+ models. Every provider.
// Smart routing by cost, latency, availability.
// Automatic failover across providers.
// Real cost tracking per token, per team.
// A gateway that actually thinks.
Why teams outgrow Cloudflare AI Gateway
Cloudflare is great for caching and basic proxying. When you need real routing, governance, and cost control, you need more.
Smart routing, not just proxying
Cloudflare AI Gateway proxies requests to a single provider. Requesty routes intelligently across 400+ models based on cost, latency, and availability. Automatic failover across providers is built in.
Real cost optimization
Cloudflare caches responses but does not optimize which model you use. Requesty combines prompt caching with smart routing to cut costs by up to 80%. You save on both the cache layer and the model selection layer.
Full model access, one key
Cloudflare AI Gateway requires you to bring your own API keys for every provider. Requesty gives you one key that works with OpenAI, Anthropic, Google, Mistral, and 30+ more providers instantly.
EU hosting for compliance
Dedicated EU endpoint in Frankfurt for GDPR compliance. Cloudflare runs on its edge network, which makes data residency guarantees harder to enforce for regulated workloads.
Enterprise governance
Five layer policy hierarchy with org, group, service account, user, and key level controls. Budget caps per team. PII masking guardrails. SOC2, GDPR, and HIPAA compliance.
Deep observability
Per model cost, TTFT, P95 latency, error rates, and session reconstruction. Cloudflare gives you request counts and basic analytics. Requesty gives you the full picture.
Feature comparison
Both can proxy LLM requests. Here is where they differ.
Switch in 2 minutes
Same OpenAI compatible format. Just change the base URL and API key.
# Before (Cloudflare AI Gateway)
base_url = "https://gateway.ai.cloudflare.com/v1/..."
api_key = "sk-..." # provider specific
# After (Requesty)
base_url = "https://router.requesty.ai/v1"
api_key = "rq-..."
# One key for every provider.
# $10 free credits. Smart routing included.