Skip to navigation
Vision APIsVision

Rate Limits

For chat over text, images and short videos, use the OpenAI-compatible Chat Completions API.

Rate limits keep usage fair and the service stable for all customers. When you exceed a limit, the API returns a 429 Too Many Requests error.

Request limits

All limits are per API key, per 24-hour rolling window.

EndpointRequests per Day
POST /v1/images/upload100
POST /v1/images/search100
POST /v1/clips10

Rate limit headers

Every response carries these usage headers:

HeaderDescription
x-ratelimit-requests-limitMaximum requests allowed in the time window
x-ratelimit-requests-remainingRequests remaining in the current window
x-ratelimit-requests-resetSeconds until the rate limit window resets
retry-afterSeconds to wait before retrying (0 if allowed, positive if rate limited)

Example

HTTP/1.1 429 Too Many Requests
x-ratelimit-requests-limit: 100
x-ratelimit-requests-remaining: 0
x-ratelimit-requests-reset: 86400
retry-after: 3600

Handling 429 errors

When you hit a rate limit:

  1. Check X-RateLimit-Reset to know when you can retry
  2. Back off exponentially between retries
  3. Batch operations where you can
{
  "error": {
    "code": "rate_limit_exceeded",
    "message": "You have exceeded your daily request limit for this endpoint."
  }
}

Higher limits

Enterprise plans have higher limits. Contact us if you need:

  • Higher request quotas
  • Increased content size limits

For pricing information, see the Vision Pricing page.