# Limits

> The categories of limits in Lorito and where to find the exact values for your plan.

Source: https://www.lorito.ai/docs/reference/limits

Lorito applies limits to keep the platform fast and fair for everyone. They exist so that normal usage is never affected. Most people won't come close to them.

> This page explains **which kinds** of limits exist. The **exact value** of each one depends on your plan and is shown in-product, so there are no fixed numbers here. To see your own limits and how much you've used, open the **Billing** area and the **Knowledge** area in your dashboard.

## Knowledge source limits

The number of sources you can add is limited, and counted **separately by source type**. Websites, PDFs, and text or markdown each have their own allowance. They're counted separately because they differ in cost and effort to index, so a website doesn't use up your room for PDFs and vice versa.

There are also limits on **how much a single source can cover**, for example how many items are indexed from one website. When a source has more content than your plan covers, Lorito indexes up to the limit and quietly notes the rest, rather than failing.

## Knowledge capacity

Beyond per-source limits, there's a ceiling on **how much knowledge you can store overall**. This applies both:

* **Per project**, the total content held in one project's knowledge base, and
* **Across your organisation**, the combined content of all your projects.

You see this as a **fill bar** in the **Knowledge** area, a simple "how full am I" indicator rather than a raw number. Lorito gives you a gentle heads-up as you approach the ceiling, so you can add a higher plan or remove unused sources before it bites.

There's also a rolling allowance on **how much content you can index over time**, which resets on your billing cycle.

## Per-agent integrations and skills

To keep your agent focused and responsive, there are limits on **how many tools an agent can have enabled at once**, across skills, webhooks, and connected integrations. A project's overall library can be larger; the limit applies to what's switched on for a specific agent.

These limits exist for quality as much as cost: an agent with too many tools enabled can lose track of which one to use, so a tighter set generally produces better answers.

## Channel and branding entitlements

Some capabilities aren't about quantity but about **whether your plan includes them at all**. These include:

* **Custom domains** for your chat and embed snippets.
* **Organisation branding.** White-labelling your dashboard and chat surfaces.
* **Chat-surface branding.** Applying your own theme and logo to a project's public chat.

When a capability isn't on your current plan, Lorito shows an upgrade prompt where you'd use it. See [Plans & entitlements](https://www.lorito.ai/docs/account-billing/plans-and-entitlements) for the full picture.

## Projects

The number of projects you can run at once is also limited by your plan. Deleting a project frees up a slot, so you can create another in its place.

## Related

* [Plans & entitlements](https://www.lorito.ai/docs/account-billing/plans-and-entitlements) explains how your plan sets features and limits
* [Credits](https://www.lorito.ai/docs/account-billing/credits) covers the separate allowance that meters your agent's usage
* [Troubleshooting](https://www.lorito.ai/docs/reference/troubleshooting) helps when something doesn't behave as expected
