---
title: How TapMind custom adapters work
description: The mental model that makes every integration page make sense.
---

> **For AI agents:** the complete documentation index is at [llms.txt](/llms.txt). Append `.md` to any page URL for its markdown version.

The one idea to hold onto: with the Custom Adapter SDKs, you do not call TapMind directly. You integrate your mediation SDK exactly as you normally would, and TapMind rides inside it as a custom network. Your mediation platform loads the TapMind adapter and asks it for an ad whenever TapMind's line item is eligible.

## The request flow

Your app → Mediation SDK (AdMob / MAX / LevelPlay / GAM)

→ picks TapMind from the waterfall

→ loads the TapMind adapter class

→ TapMind backend returns an ad

→ ad renders in your app

## What this means in practice

- You add TapMind in two places: a dependency in your build, and a custom-event/custom-network entry in your mediation dashboard.
- After that, you request ads through your mediation SDK's normal API — there is no separate TapMind "load ad" call (the Orchestration SDK is the exception; it has its own API).
- TapMind only serves when your waterfall reaches its line item, at the eCPM you configure.

## Key terms

| Term | Meaning |
|------|---------|
| Custom event / custom network | The mechanism your mediation platform uses to call a third-party adapter like TapMind. |
| Class name | The exact adapter class your mediation dashboard loads. Must match the published SDK exactly. |
| placementName | A per-placement identifier you enter in the dashboard. Provided by your account manager. Case-sensitive, exact-match. |
| Network key | A per-account key that identifies TapMind in Unity LevelPlay. |
| eCPM / rate | The manual price that sets TapMind's position in your waterfall. |

<Callout type="info">
This page is the prerequisite mental model for every integration page. If an ad isn't serving, 90% of the time it traces back to one of the four bolded values above not matching exactly.
</Callout>