Rate Limits

Sawalni enforces rate limits to ensure fair usage and system stability.

Default Limits

Tier	Requests/min	Tokens/min
Free	60	40,000
Pro	300	200,000
Enterprise	Custom	Custom

Rate Limit Headers

Every API response includes rate limit information:

text

X-RateLimit-Limit: 60
X-RateLimit-Remaining: 45
X-RateLimit-Reset: 1704556800

Handling 429 Errors

When you exceed your limit, you'll receive a 429 Too Many Requests response:

json

{
  "error": {
    "message": "Rate limit exceeded",
    "type": "rate_limit_error",
    "code": "rate_limit_exceeded"
  }
}

Best Practices

Implement exponential backoff — Wait 1s, 2s, 4s, 8s… between retries
Cache responses when the same input produces deterministic output
Batch requests — Use array inputs for embeddings
Monitor headers — Check X-RateLimit-Remaining proactively
Upgrade your tier if you consistently hit limits

Example: Retry with Backoff

python

import time
from openai import OpenAI, RateLimitError

client = OpenAI(
    api_key="YOUR_API_KEY",
    base_url="https://api.sawalni.com/v1"
)

def chat_with_retry(messages, max_retries=3):
    for attempt in range(max_retries):
        try:
            return client.chat.completions.create(
                model="sawalni-beta",
                messages=messages
            )
        except RateLimitError:
            wait = 2 ** attempt
            print(f"Rate limited, waiting {wait}s...")
            time.sleep(wait)
    raise Exception("Max retries exceeded")