Overview
Rate limits are restrictions on the number of API requests you can make within a specific time window. They ensure fair usage, maintain service quality, and prevent abuse.
Rate Limit Reset
Rate limits are tracked using sliding windows. When you hit a limit, the API will return a 429 Too Many Requests status with details about when the limit resets.
Global Rate Limits
These limits apply to all API endpoints and all models. All plans are now subject to the same base rate limits.
| Endpoint | Limit | Window | Description |
|---|---|---|---|
| /v1/generate | 5 requests | 10 minutes | Applies to all video generation requests (all plans) |
| /v1/status | No limit | - | Status checks are unlimited for polling |
| /v1/generations | No limit | - | Listing your generations is unlimited |
| /v1/user | No limit | - | User information endpoint is unlimited |
Why These Limits?
All plans are now limited to 5 requests per 10 minutes on /v1/generate to ensure fair usage and maintain service quality. Some models have additional specific limits (see below).
Model-Specific Rate Limits
Some AI models have additional rate limits due to their computational cost or third-party provider restrictions.
| Model | Model Key | Limit | Window | Applies To |
|---|---|---|---|---|
| Kling 3 | kling-3 | 5 requests | 15 minutes | All plans |
| Seedance 2 | seedance-2 | 5 requests | 15 minutes | All plans |
| Sora 2 | sora_2 | 50 (Standard) / 100 (Unlimited) | 1 hour | Plan-dependent |
| All other models | * | No additional limit | - | - |
Kling 3 & Seedance 2 Restrictions
Les modèles Kling 3 et Seedance 2 ont une limite stricte de 5 générations par 15 minutes pour tous les plans :
- Coûts computationnels très élevés
- Qualité vidéo premium
- Limitations des fournisseurs tiers
- S'applique même aux plans Unlimited
Sora 2 Special Restrictions
Le modèle Sora 2 a une limite horaire variable selon le plan :
- 50 générations/heure pour les plans Standard
- 100 générations/heure pour les plans Unlimited
- Coûts computationnels élevés
- Capacités AI avancées
Plan Tiers & Limits
Your plan tier determines your monthly generation quota. All plans now have uniform base rate limits with model-specific restrictions.
| Plan | Monthly Quota | Global Rate Limit | Kling 3 / Seedance 2 | Sora 2 Limit | Price |
|---|---|---|---|---|---|
| Free | 10 generations | 5 per 10min | 5 per 15min | 50 per hour | $0 |
| Starter | 100 generations | 5 per 10min | 5 per 15min | 50 per hour | $29/month |
| Pro | 500 generations | 5 per 10min | 5 per 15min | 50 per hour | $99/month |
| Unlimited Premium | Unlimited | 5 per 10min | 5 per 15min | 100 per hour | $199/month |
bolt Rate Limit Summary
- check_circle 5 requêtes/10min limite globale pour tous les plans
- check_circle 5 requêtes/15min pour Kling 3 et Seedance 2
- check_circle Status checking is always unlimited
upgrade Unlimited Plan Benefits
- star Unlimited monthly generations quota
- star 100 req/hour for Sora 2 (vs 50 for Standard)
- star Priority support and early access to new models
Best Practices
schedule Implement Request Queuing
Instead of firing all requests at once, queue them and space them out to stay within rate limits.
// Wait 1 second between requests
await sleep(1000);
cached Use Exponential Backoff
When you receive a 429 error, wait progressively longer before retrying.
// Retry with exponential backoff
await sleep(2^retryCount * 1000);
monitor_heart Monitor Rate Limit Headers
The API response includes rate limit information in the rate_limit object:
"rate_limit": {
"current_usage": 15,
"limit": 50,
"seconds_until_reset": 847
}
batch_prediction Batch Your Operations
Group related generation requests and space them intelligently. For Kling 3 and Seedance 2, plan your 5 generations per 15 minutes strategically. For Sora 2, leverage the higher hourly limits (50-100/hour) for high-volume use cases.
Handling Rate Limit Errors
When you exceed a rate limit, the API returns a 429 Too Many Requests error with detailed information.
Error Response Example
{
"error": "Rate limit exceeded",
"message": "Too many requests. Please wait before trying again.",
"details": {
"current_usage": 5,
"limit": 5,
"window": "10 minutes",
"reset_time": "2025-10-09 15:30:00",
"seconds_until_reset": 312
}
}Model-Specific Rate Limit Error (Sora 2)
{
"error": "Sora-2 rate limit exceeded",
"message": "The Sora-2 model is limited to 5 generations per hour for the Unlimited plan. Contact support for higher limits.",
"details": {
"current_usage": 10,
"limit": 10,
"window": "1 hour",
"reset_time": "2025-10-09 15:00:00",
"seconds_until_reset": 3247,
"upgrade_required": true,
"upgrade_plan": "Unlimited $199/month"
}
}Important Notes
- Always check the
seconds_until_resetfield to know when to retry - Respect the rate limits to avoid temporary API access suspension
- Consider upgrading your plan if you consistently hit rate limits