RLaaS Documentation
Attach rate limits to any HTTP endpoint with a single POST call. Four algorithms (fixed, sliding, token, leaky), fast rule edits, and a minimal allow/deny response model.
POST /check→ send{api_key, endpoint, key}— receive allow or 429.- Strategies: fixed window, sliding window, token bucket, leaky bucket.
fail_openflag lets you pick strict denial vs availability on internal errors.
<endpoint, strategy, key_by, limit, window, fail_open>. Sliding smooths window edges; token & leaky enforce steady rates; fixed is the lowest-overhead baseline. Benchmarks show fixed fastest; sliding balances smoothness vs cost. Treat any response without "allowed": true as a denial for consistent enforcement.