Getting Started

1. Create an account

2. Complete onboarding

Add your platform details and select the ad formats you want to use.

3. Create an API key

Create an API key in the AdMesh dashboard and use it as apiKey in the integration below.

4. Install the frontend SDK

npm install admesh-ui-sdk@latest

For Flutter, follow the Flutter integration steps.

5. Wrap your chat with AdMeshProvider

Use your AdMesh API key and keep the same sessionId throughout the conversation. Pass previous messages through the optional messages prop to help AdMesh recommend products that match the user’s earlier preferences, budget, and needs.

import { AdMeshProvider } from 'admesh-ui-sdk';

// Use previous turns from your application's chat state, oldest first.
const previousMessages = [
  { role: 'user', content: 'I need a laptop for coding.' },
  { role: 'assistant', content: 'What is your budget?' },
  { role: 'user', content: 'Under $1,500, with good battery life.' },
];

<AdMeshProvider
  apiKey={apiKey}
  sessionId={sessionId}
  messages={previousMessages} // Optional: conversation history for more relevant recommendations
>
  <YourChat />
</AdMeshProvider>

You can omit messages if conversation history is unavailable. Recommendations still work using the current query. Each message contains role and content strings. Keep this array updated from your application’s chat state as the conversation progresses; the provider does not collect previous messages automatically. For example, when the user next asks “Which one would you recommend?”, the previous messages give AdMesh the laptop requirements and budget behind that question. Continue passing the current user query separately as query to the recommendation component in the next step. AdMeshRecommendations and WeaveAdFormatContainer include this history in their frontend recommendation requests. For Weave recommendations fetched by your backend, pass conversation history to that backend integration as well.

6. Integrate your selected format

Tail, Product & Bridge
Weave (optional)

Add AdMeshRecommendations below each assistant response. Pass the original user query and its user message ID.

import { AdMeshRecommendations } from 'admesh-ui-sdk';

<AdMeshRecommendations
  messageId={userMessageId}
  query={userQuery}
/>

Use the formats selected in your dashboard; no format prop is needed. For Bridge’s Talk to Agent action, connect the delegation callbacks.

Complete these additional steps only if you selected Weave.

1. Install a backend SDK

npm install @admesh/weave-node@latest

pip install admesh-weave-python

2. Fetch recommendations before generating the response

Follow the Node.js or Python setup. Pass the returned recommendation context into your LLM and preserve AdMesh tracking links in its output.

3. Wrap the assistant response

Use this container in place of the standalone AdMeshRecommendations unit for the same response. Pass the assistant message ID returned by your backend.

import { WeaveAdFormatContainer } from 'admesh-ui-sdk';

<WeaveAdFormatContainer
  messageId={assistantMessageId}
  query={userQuery}
  fallbackFormat="tail"
>
  <YourRenderedAssistantResponse />
</WeaveAdFormatContainer>

4. Connect streaming events

Dispatch these events from your chat’s streaming handlers, using the same assistant message ID and provider session ID.

import {
  dispatchStreamingStartEvent,
  dispatchStreamingCompleteEvent,
} from 'admesh-ui-sdk';

// When streaming starts and the assistant message ID is available:
dispatchStreamingStartEvent(assistantMessageId, sessionId);

// When streaming finishes and the response has rendered:
dispatchStreamingCompleteEvent(assistantMessageId, sessionId);

When testing, check that woven links receive an Ad label. Test a response without woven links to verify fallback recommendations when an eligible ad is available.

7. Test your integration

Open your app and send a query. Verify that an eligible recommendation renders and that its links work.

Advertisers

AI Platform

Publishers

SDKs

Core Concepts

Policies

FAQ

1. Create an account

2. Complete onboarding

3. Create an API key

4. Install the frontend SDK

5. Wrap your chat with AdMeshProvider

6. Integrate your selected format

1. Install a backend SDK

2. Fetch recommendations before generating the response

3. Wrap the assistant response

4. Connect streaming events

7. Test your integration

​1. Create an account

​2. Complete onboarding

​3. Create an API key

​4. Install the frontend SDK

​5. Wrap your chat with AdMeshProvider

​6. Integrate your selected format

​1. Install a backend SDK

​2. Fetch recommendations before generating the response

​3. Wrap the assistant response

​4. Connect streaming events

​7. Test your integration

1. Create an account

2. Complete onboarding

3. Create an API key

4. Install the frontend SDK

5. Wrap your chat with AdMeshProvider

6. Integrate your selected format

1. Install a backend SDK

2. Fetch recommendations before generating the response

3. Wrap the assistant response

4. Connect streaming events

7. Test your integration