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

# What are TTAPI accounts and queue system? How do API concurrency and task queues work?

> Learn how TTAPI accounts and queue system work, including API concurrency limits, request queuing, queue priority, and rate limiting.

## What queue means on TTAPI

On TTAPI, `queue` does not mean traditional rate limiting such as `QPS` or `RPM`. It means the number of **in-progress tasks** a product can hold for your account at the same time.

An in-progress task usually means one of these:

* The task has been accepted but is still waiting in queue
* The task is already running but has not finished yet
* The final result must be fetched later through polling or webhook callbacks

So queue is closer to concurrent active jobs than per-second request throughput.

## Which products use queue limits

| Product      | Queue limit | Why                                                  |
| :----------- | :---------- | :--------------------------------------------------- |
| `Midjourney` | Yes         | Jobs take time and results are usually asynchronous  |
| `Suno`       | Yes         | Music generation is long-running and asynchronous    |
| `Sora`       | Yes         | Video jobs are resource-heavy and slower to complete |
| `Producer`   | Yes         | Audio generation and processing are not instant jobs |

These products all share the same pattern: longer execution time, higher resource usage, and delayed final results.

## Why queue limits exist

TTAPI relies on a finite account pool and execution capacity behind each product line.

Without queue controls, a single customer could suddenly push a large amount of unannounced concurrency and fill the shared account pool, which can lead to:

* One product line being saturated
* Other users being blocked from submitting or fetching tasks
* Large queue buildup across the service
* Upstream provider risk controls or service degradation

The goal of queue limits is not to block normal usage. It is to prevent abusive or unplanned traffic spikes from destabilizing the platform.

## Queue expansion

If you expect stable and predictable traffic, you can [contact us](/grids/en/about/connect) to request more queue capacity.

Useful details to include:

* Which product needs more queue
* Expected peak concurrent jobs
* Typical daily task volume
* Time window for the traffic increase
* Whether the increase is tied to a launch, campaign, or scheduled growth

Queue expansion is currently **free**, but we may include it in future value-added service plans.