Skip to main content
Usage Tiers are trust-based rate-limit packages for the Visor Public API. They are separate from pricing: request prices come from the usage class shown in X-Usage-Class and the pricing version shown in X-Pricing-Version.

​
Account-level limits

Rate limits are enforced per API account, not per API key. All keys on the same account share the same limits.
Usage TierRequests / 10 secondsRequests / 60 seconds
Tier 12060
Tier 2100100
Tier 3200600
Tier 45003000
All API accounts start in Tier 1. Your dashboard is the source of truth for your current effective limits, including any account-specific override.

​
Tier movement

Higher tiers are designed to limit abuse and protect platform capacity while trusted accounts ramp up. Automatic promotion uses verified paid API credits, days since the first successful paid credit, and current good standing. Promotional or manually granted beta credits do not count toward automatic promotion.
Target TierVerified paid API creditsDays since first successful paid creditReview
Tier 2$257Automatic when in good standing
Tier 3$15014Automatic when in good standing
Tier 4$50030Manual review required
Good standing means the API account is active, billing state is active or low balance, available balance is positive, and there has been no recent chargeback reversal. Tier movement is upward by default; lower limits are an operational action for abuse, payment risk, customer request, or support intervention.

​
Account-specific overrides

Some accounts may have custom effective limits that differ from the public tier table. Overrides can raise limits for trusted partners or lower limits for abuse mitigation. If your dashboard differs from this guide, use the dashboard value.

​
Rate-limit responses

When you exceed your current effective limit, the API returns HTTP 429: 
{
  "error": {
    "type": "rate_limit_error",
    "code": "rate_limit_exceeded",
    "message": "Rate limit exceeded."
  },
  "_tag": "PublicApiRateLimitError"
}
Rate-limited responses include Retry-After when available. Wait at least that many seconds before retrying.

​
Integration guidance

Use client-side backoff for 429 responses and avoid sending parallel bursts beyond your tier. For scheduled jobs, spread requests across the minute instead of sending them all at once.