> ## Documentation Index
> Fetch the complete documentation index at: https://docs.somya.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Rate limits & quota

> Credit-based quota, usage tracking, and rate limiting.

## Credit-based quota

Usage is metered in **credits**, allocated **per billing cycle** and tracked
separately for TTS and ASR. Your allocation depends on your
[plan](/guides/pricing):

| Plan | TTS credits / cycle | ASR credits / cycle |
| ---- | ------------------- | ------------------- |
| Free | 1,000               | 1,000               |
| Pro  | 10,000              | 10,000              |

When you exhaust a cycle's credits, requests return **`402`** until the cycle
resets or you upgrade.

## Checking usage

Query your current usage with:

```bash theme={null}
curl https://api.somya.ai/v1/usage -H "X-API-Key: YOUR_API_KEY"
```

Check this proactively (e.g. before large batches) so you can upgrade or throttle
before hitting `402`.

## Rate limiting

Beyond quota, the API rate-limits bursts of requests. If you send too many too
quickly you'll get **`429`**:

```json theme={null}
{ "success": false, "data": null, "error": { "code": "RATE_LIMITED" } }
```

Handle it with exponential backoff (see [Errors](/guides/errors#retries)). For
high throughput, queue requests and cap concurrency rather than firing them all
at once.

<Note>
  Exact per-second/per-minute rate-limit thresholds depend on the deployment and
  plan. Design for backoff on `429` rather than a fixed request rate.
</Note>