> For the complete documentation index, see [llms.txt](https://docs.qolaba.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.qolaba.ai/video-generation/output-controls/number-of-generations.md).

# Number of Generations

Number of Generations controls how many video outputs Qolaba produces from a single prompt and settings configuration in one run. Generating multiple outputs at once gives you variations to compare — without re-entering your settings each time.

***

#### How It Works

Each generation produces one unique video output from the same prompt, reference media, and output settings. Increasing the count produces multiple variations in a single run.

Credit cost scales linearly with generation count:

> **Total credits = credits per video × number of generations**

**Example:** If a single generation costs 500 credits:

<table><thead><tr><th width="277.0390625">Generations</th><th>Total Credits</th></tr></thead><tbody><tr><td>1</td><td>500</td></tr><tr><td>2</td><td>1,000</td></tr><tr><td>3</td><td>1,500</td></tr></tbody></table>

The total credit cost for your current configuration is displayed on the **Generate button** before you confirm — so you always know what will be consumed before committing.

***

#### When to Generate Multiple vs. One at a Time

**Generate multiple when:**

* Your prompt and settings are already validated from a previous run
* You want variations to compare before selecting a final output
* You are preparing a client presentation with options
* You are running a confirmed final production batch

**Generate one at a time when:**

* You are working with a new or untested prompt
* You are trying a model or settings combination for the first time
* You want to validate output direction before spending credits on a full batch
* You are making incremental prompt refinements between runs

{% hint style="info" %}
Always run a single generation first when working with a new prompt or model. Refine based on that result before scaling to multiple generations — this is the most effective way to avoid spending credits on outputs that need significant rework.
{% endhint %}

***

#### A/B Generation

Setting generations to **2** is particularly useful for A/B comparison — two variations of the same prompt are produced side-by-side, giving you a direct quality and style comparison from a single run. This is more credit-efficient than running two separate single generations when your settings are already confirmed.