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

# Concurrency - Max Simultaneous Candidates

> Understand AutoProctor's concurrency limits that control how many candidates can start tests simultaneously within a 60-second window.

Concurrency controls how many candidates can start any test under your account within the same 60-second window. Understanding these limits helps you plan large-scale test sessions and avoid unexpected delays for candidates.

## What Is Max Concurrency?

Max concurrency is the maximum number of candidates who can begin a test within any rolling 60-second period. This limit applies across **all tests on your account**, not per individual test.

<Note>
  If you have three active tests and a concurrency limit of 1,000, the 1,000 limit is shared across all three tests combined -- not 1,000 per test.
</Note>

## Concurrency Limits by Plan

### Tests with Proctoring Enabled

| Plan     | Max Concurrency         |
| -------- | ----------------------- |
| Standard | 1,000 candidates/minute |
| Premium  | 2,000 candidates/minute |
| Elite    | 5,000 candidates/minute |

### Tests Without Proctoring

| Plan     | Max Concurrency       |
| -------- | --------------------- |
| Standard | 50 candidates/minute  |
| Premium  | 100 candidates/minute |
| Elite    | 150 candidates/minute |

## Automated Staggering

When demand exceeds your plan's capacity, AutoProctor automatically staggers candidate entry. Candidates who exceed the limit see a button with a countdown of up to 60 seconds before they can access the test.

<Note>
  If your test has a duration configured, AutoProctor automatically adjusts each candidate's timer to account for staggering delays. No candidate loses test time due to concurrency limits.
</Note>

## SDK Users

If you use AutoProctor's JavaScript SDK on your own website, you receive a default limit of 5,000 candidates per minute. When this limit is exceeded, the SDK returns a 429 (Too Many Requests) response. You must handle rate-limiting independently in your implementation.

<Warning>
  SDK users do not benefit from automatic staggering. You must build your own retry logic to handle 429 responses when your concurrency limit is exceeded.
</Warning>

## Increasing Your Limits

If you need higher concurrency limits, [contact the AutoProctor team](/contact-us):

| Increase Type              | Approximate Cost                                  | Details                    |
| -------------------------- | ------------------------------------------------- | -------------------------- |
| Temporary (one-time event) | \$500 per 500 candidates/minute increment         | One-time fee               |
| Permanent                  | \$2,500/month per 500 candidates/minute increment | Annual commitment required |

<Note>
  The platform supports up to a maximum of 25,000 candidates per minute.
</Note>

## Related Resources

* [Timer Settings](/timer-settings) -- Configure test duration and time windows
* [Payments and Credits](/payments-and-credits) -- Plan details and pricing information
* [Best Practices for Test Creators](/best-practices-for-exam-creators) -- Tips for smooth test administration
* [Before You Get Started](/before-you-get-started) -- Critical requirements for using AutoProctor
* [Contact Us](/contact-us) -- Reach the AutoProctor team for limit increases