API Key Limits

API Key is your credential to access ai.TokenHub services. Understanding its limits and management is important for stable service usage.

API Key Types

Type	Use Case	Features
Test Key	Development	Lower rate limits, limited free quota
Production Key	Production	Higher rate limits and quota

Rate Limits

Rate limits are set to protect service stability:

Limit Type	Description
RPM (Requests Per Minute)	Maximum requests per minute
TPM (Tokens Per Minute)	Maximum tokens per minute
Concurrent requests	Maximum simultaneous requests

Different models and packages have different rate limits. Check the console for specific values.

Limit Exceeded Response

When exceeding rate limits, API returns 429 error:

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

Best Practices

1. Implement Retry Mechanism

import time
from openai import OpenAI, RateLimitError

client = OpenAI(
    base_url="https://ai-tokenhub.com/v1",
    api_key="YOUR_API_KEY"
)

def call_with_retry(messages, max_retries=3):
    for i in range(max_retries):
        try:
            return client.chat.completions.create(
                model="gpt-4o",
                messages=messages
            )
        except RateLimitError:
            wait_time = (i + 1) * 2
            time.sleep(wait_time)
    raise Exception("Max retries exceeded")

2. Batch Request Optimization

Combine multiple small requests into batch requests:

# Not recommended: multiple individual requests
for text in texts:
    response = client.embeddings.create(input=text, model="text-embedding-3-small")

# Recommended: single batch request
response = client.embeddings.create(input=texts, model="text-embedding-3-small")

3. Monitor Usage

Regularly check API Key usage to avoid exceeding quota:

curl https://ai-tokenhub.com/v1/credits \
  -H "Authorization: Bearer YOUR_API_KEY"

Quota Management

• Each API Key has independent quota
• Charged by token usage
• Set quota warning threshold in console
• API returns insufficient_quota error when quota exhausted

Security Recommendations

1. Don't hardcode API Key: Use environment variables
2. Rotate regularly: Monthly rotation recommended
3. Separate usage: Use different Keys for dev and production
4. Monitor anomalies: Watch for abnormal request patterns