---
title: Request limits
description: Handle request, upload, and concurrency limits reliably.
---

Standard accounts have **60 requests per minute** and **5 concurrent requests**, shared across all API keys and the dashboard. Speech and voice cloning share the concurrency limit.

The per-minute allowance refills steadily with a small burst allowance. Spread requests over time rather than sending the entire minute's allowance at once. [Contact us](/contact) for higher limits.

`GET /v1/me` returns your effective limits; `null` means no account-level cap. Limit changes apply within 60 seconds. Available service capacity still applies.

## Handling a limit

On `429`, wait at least the time in `Retry-After`, then retry with bounded backoff and jitter. For concurrency limits, wait for an existing request to finish.

Operation status and cancellation have a separate allowance of 600 requests/minute, so you can still inspect or stop work after reaching the regular limit.

## Request and account limits

| Request | Limit |
| --- | --- |
| Speech | 10,000 characters; 400 per word; JSON body up to 65,536 bytes. |
| Voice cloning | 35 MiB total upload; one audio file of at most 20 MiB. We use the first 30 seconds. See [form fields](/docs/voice-cloning). |
| Clone uploads in progress | 2 per account, 8 across the service. |
| Active API keys | 20 per account. |
| Voice collections | 100 per account. |

## Website demo

The free demo allows 10 requests/minute, 3 concurrent generations, and 300 characters per request. Visitors sharing a network address share these limits.