FAIRVISORFairvisor vs. AWS API Gateway Throttling
The Situation
AWS API Gateway throttling is easy to use inside AWS and works well for standard request-rate limits.
Fairvisor is designed for AI-aware enforcement: token and cost budgets, loop protection, and policy dimensions beyond API keys.
Comparison
| Capability | Fairvisor | AWS API Gateway |
|---|---|---|
| Rate limiting | Token bucket + cost-based + token-based (TPM) | Token bucket (requests only) |
| Limit keys | Any JWT claim, header, path, User-Agent | API key, usage plan |
| Cost-based budgets | Yes, per-org, per-user, per-endpoint | No |
| AI-specific features | Loop detection, token counting, circuit breaker | No |
| AI crawler protection | User-Agent detection + rate limiting | No |
| Multi-cloud | Runs anywhere | AWS only |
| Latency profile | In-process, low-latency decision path | Managed gateway policy path |
| Shadow mode | Yes | No |
| Analytics | Built-in dashboard, CSV export | CloudWatch raw metrics |
| Pricing model | OSS edge + optional SaaS plan | Managed service pricing by AWS plan/usage |
| Vendor lock-in | None | AWS ecosystem |
| Offline operation | Yes | No |
When to Use Fairvisor
- You need AI-aware rate limiting and cost-based budgets.
- You need multi-cloud or self-hosted deployment flexibility.
- You want to avoid per-request pricing at high volume.
When to Use AWS API Gateway
- You are all-in on AWS.
- You need managed gateway features (auth, transformations, WebSocket).
- Request-based throttling is enough for your use case.
Use Them Together
- Use AWS API Gateway for routing and auth.
- Add Fairvisor as policy decision layer for advanced enforcement.
- Keep AWS convenience while adding AI-specific protection controls.