Rate Limits & Credits
Our API operates on a high-performance Pay-As-You-Go (PAYG) credit system. There are no monthly subscription tiers—you only pay for what you use.
Rate Limits
To protect our infrastructure and ensure fair usage, we enforce rate limits on a per-endpoint basis. When you exceed your limit, the API will return a 429 RATE_LIMIT_EXCEEDED error.
| Account Type | Default Limit | Custom Limit |
|---|---|---|
| Standard Account | 60 Requests / minute (per endpoint) | Not Available |
| Enterprise / High-Volume | Custom (e.g., 1,000+ RPM) | Contact Support |
We utilize a highly optimized in-memory File Cache for rate limit tracking. This guarantees sub-millisecond latency for API key validation without hitting the database on every request.
Credits & Billing Flow
We use a dynamic Pre-flight & Post-flight billing model to ensure you are only charged for exact token usage, down to the decimal.
1. Pre-Flight Check
Before executing the request, we check if you have the minimum required base_credits for the endpoint. If not, a 402 INSUFFICIENT_CREDITS error is thrown.
2. Execution
The AI model processes your request. Since LLM output length is unpredictable, the final cost cannot be determined until generation is complete.
3. Post-Flight Deduction
Once the response is generated, we calculate the exact cost based on Input/Output tokens and perform an Atomic Cache Decrement to instantly update your live balance.
4. Zero Latency
All billing deductions happen asynchronously. Your API response is sent back instantly without waiting for database ledger updates.
Endpoint Base Credits
Below is a list of the base credits required to initiate a request for each specific endpoint.
Note on Final Cost: The total credits deducted will be calculated as:
Final Cost = Base Credit + Additional Credit (based on content size/tokens generated).
| Endpoint Category | API Endpoint | Base Credits (Pre-Flight) |
|---|---|---|
| NLP | /api/v1/ai/intent-detection |
1 Credit(s) |
| NLP | /api/v1/ai/translation |
1 Credit(s) |
| Conversational | /api/v1/ai/chat |
2 Credit(s) |
| NLP | /api/v1/ai/email-reply |
2 Credit(s) |
| Content | /api/v1/ai/content-generation |
3 Credit(s) |
| Vision | /api/v1/ai/image-generation |
5 Credit(s) |
| Account | /api/v1/account/usage |
0 Credit(s) |
| Account | /api/v1/account/usage-summary |
0 Credit(s) |
| Account | /api/v1/account/api-key/current |
0 Credit(s) |
| NLP | /api/v1/ai/json-extractor |
3 Credit(s) |
| NLP | /api/v1/ai/email-classification |
2 Credit(s) |
| NLP | /api/v1/ai/sentiment-analysis |
1 Credit(s) |
| Security | /api/v1/ai/spam-detection |
2 Credit(s) |
| NLP | /api/v1/ai/text-summarize |
3 Credit(s) |
| NLP | /api/v1/ai/grammar-fix |
1 Credit(s) |
| NLP | /api/v1/ai/pii-redaction |
2 Credit(s) |
Processing Cost Calculation
Because AI processing requires more compute power for larger texts, we add a Processing Cost based on the character length of your input.
| Input Size (Characters) | Processing Cost (Additional Credits) |
|---|---|
| Up to 500 chars (Small) | +1 Credit |
| 501 to 2,000 chars (Medium) | +2 Credits |
| 2,001 to 5,000 chars (Large) | +4 Credits |
| 5,001+ chars (Extra Large) | +4 Credits plus +1 Credit per 1,000 additional chars |
If you call an endpoint with a Base Credit of 5, and send 1,200 characters of text:
Base (5) + Processing (2) = 7 Credits deducted.Note: The absolute minimum charge for any request is 2 credits.
Because final credit costs are calculated after generation, it is possible for a heavy request to push your account into a small negative balance. If your balance drops below zero, subsequent pre-flight checks will fail until you recharge.