🐢 Rate Limit - API Resilience
Fair usage policies protecting platform stability and ensuring equitable resource access
Our rate limits are designed to enable robust integrations while maintaining service reliability for all users. These guardrails ensure your application can scale sustainably within our ecosystem.
Service Account Rate Limits
All authenticated API requests made by service accounts are subject to rate limiting based on client identity.
Current Limits
| Limit Type | Quota | Window | Sustained Rate |
|---|---|---|---|
| Service Account | 4,500 requests | 10 minutes | 7.5 requests/second |
How it works:
- Each service account has an independent quota tracked by
client_id - Requests consume tokens from a bucket that continuously refills
- Burst capacity allows up to 4,500 immediate requests
- Tokens replenish at 7 per second for sustained usage
Understanding Rate Limit Headers
Service account requests (authenticated via JWT without user claims) include rate limit information in response headers. Regular user requests and unauthenticated requests do not include these headers.
x-ratelimit-remaining: 4499
x-ratelimit-replenish-rate: 7
x-ratelimit-burst-capacity: 4500
x-ratelimit-requested-tokens: 1
| Header | Description |
|---|---|
x-ratelimit-remaining |
Tokens remaining after this request (decreases with each request, refills over time) |
x-ratelimit-replenish-rate |
Tokens added per second (7) |
x-ratelimit-burst-capacity |
Maximum tokens in bucket (4500) |
x-ratelimit-requested-tokens |
Tokens consumed per request (1) |
When rate limited (429 response):
HTTP/1.1 429 Too Many Requests
Retry-After: 1
X-RateLimit-Remaining: 0
The Retry-After header indicates seconds to wait before retrying.
Recovery:
- Immediate: New tokens available within ~133ms
- Full capacity: Restored after 10 minutes of reduced activity
- No blocking: Your service account remains active, only rate-limited
Best practices:
- Implement exponential backoff when receiving 429 responses
- Cache responses where possible to reduce request volume
- Monitor your rate limit headers to stay within quota
- Distribute requests evenly rather than bursting
FAQ
Are webhooks counted toward rate limits?
No, incoming webhooks from LIKE MAGIC do not consume your rate limit quota
What happens if I exceed limits repeatedly?
Persistent abuse may trigger additional security reviews, but normal bursts within backoff strategies are expected behavior.
Do different environments have different limits?
Rate limits are consistent across environments to ensure realistic testing and predictable production behavior.
Can I check my current rate limit status?
Yes, service account requests include x-ratelimit-remaining, x-ratelimit-replenish-rate and x-ratelimit-burst-capacity headers in every response. Watch x-ratelimit-remaining to see your current token count.
Build sustainably. Scale confidently. Respect the limits 🎯